<div dir="ltr"><div style="font-size:12.8px"><div><div class="gmail_default" style="font-size:12.8px"><font color="#000000" face="arial, helvetica, sans-serif">When: Tuesday, March 14th at 11:00 am </font></div><div class="gmail_default" style="font-size:12.8px"><font color="#000000" face="arial, helvetica, sans-serif"><br></font></div><div class="gmail_default" style="font-size:12.8px"><font color="#000000" face="arial, helvetica, sans-serif">Where: <span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_8355097314499745974gmail-m_8816702620734345313gmail-m_3262824545120381495gmail-m_-1141671822915777344gmail-m_-7219251726624328345gmail-m_-8588148075564318222gmail-m_-8767966813928691312gmail-m_-1542318334608687154gmail-m_5717104778280916634gmail-m_4845490158781220632gmail-m_5124567205141626540gmail-m_3209361100497750746gmail-m_2953668934074478317gmail-m_-3155518689668024534m_9067904842688472155gmail-m_3071693547520408192gmail-il">TTIC</span>, 6045 S Kenwood Avenue, 5th Floor, Room 526</font></div><div class="gmail_default" style="font-size:12.8px"><font color="#000000" face="arial, helvetica, sans-serif"><br></font></div><div><span style="font-size:12.8px"><font color="#000000" face="arial, helvetica, sans-serif">Who: </font></span><font color="#000000" face="arial, helvetica, sans-serif">Nan Jiang, University of Michigan</font></div><div style="font-size:12.8px"><span style="font-size:12.8px"><font color="#000000" face="arial, helvetica, sans-serif"><br></font></span></div></div><span style="font-size:12.8px"></span></div><div style="font-size:12.8px"><br></div><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px">Title:<span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span></span><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">New</span></span><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> R</span></span><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">esults</span></span><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span></span><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">in</span></span><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> S</span></span><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px">tatistical</span><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> R</span></span><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px"><wbr>einforcement</span></span><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> L</span></span><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">earning</span></span><br class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><br class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px">Abstract:</span><br class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px">Recently,<span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span></span><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">reinforcement</span></span><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span></span><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px">learni<wbr>ng</span><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span>(RL) has achieved inspiring success<span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span></span><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">in</span></span><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span>game playing domains, including human-level control<span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span></span><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">in</span></span><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span>Atari games and mastering the game of Go. Looking into the future, we expect to build machine<span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span></span><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">learning</span></span><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span>systems that use RL to turn predictions into actions; applications include robotics, dialog systems, online education, adaptive medical treatment, to name but a few.<span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span></span><br class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><br class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">In</span></span><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span>this talk, I show how theoretical insights from supervised<span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span></span><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">learning</span></span><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span>can help understand RL, and better appreciate the unique challenges that arise from multi-stage decision making. The first part of the talk focuses on an interesting phenomenon, that a short planning horizon can produce better policies when there is limited data. I explain it by making a formal analogy to empirical risk minimization, and argue that a short planning horizon helps avoid overfitting. The second part of the talk concerns a core algorithmic challenge<span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span></span><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">in</span></span><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span>state-of-the-art RL: sample-efficient exploration<span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span></span><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">in</span></span><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span>large state spaces. I introduce a<span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span></span><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">new</span></span><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span>complexity measure, the Bellman rank, which allows us to apply a unified algorithm to a number of important RL settings,<span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span></span><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">in</span></span><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span>some cases obtaining polynomial sample complexity for the first time.</span><br class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><br class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px">Bio:</span><br class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-il">Nan</span> <span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-il">Jiang</span> is a PhD candidate<span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span></span><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">in</span></span><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span>Computer Science and Engineering at University of Michigan, He works with Satinder Singh on a variety of topics related to<span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span></span><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">reinforcement</span> </span><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"></span><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">learning</span></span><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px">. Specific research interests include provable use of function approximation, off-policy evaluation, state representation<span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span></span><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">learning</span></span><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px">, spectral<span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span></span><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">learning</span></span><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span>of dynamical systems, and inverse RL for AI safety. <span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-il">Nan</span> received his bachelor degree<span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span></span><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">in</span></span><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span>Control and Automation from Tsinghua University<span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span></span><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">in</span></span><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span>2011. He received the Best Paper Award at AAMAS 2015, and Rackham Predoctoral Fellowship<span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span></span><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">in</span></span><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space gmail-m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span>2016.</span><div style="font-size:12.8px"><font color="#212121" face="sans-serif"><br></font></div><div style="font-size:12.8px"><font color="#212121" face="sans-serif"><br></font></div><div style="font-size:12.8px"><font color="#212121" face="sans-serif">Host: <a href="mailto:mwalter@ttic.edu" target="_blank">Matthew Walter</a></font></div><div style="font-size:12.8px"><br></div><div style="font-size:12.8px"><br></div><div style="font-size:12.8px"><br></div><div style="font-size:12.8px"><br></div><div style="font-size:12.8px"><br></div><div class="gmail_extra"><br clear="all"><div><div class="gmail_signature" data-smartmail="gmail_signature"><div dir="ltr"><div><div dir="ltr"><div><div dir="ltr"><div><div dir="ltr"><font face="arial, helvetica, sans-serif">Mary C. Marre</font><div><font face="arial, helvetica, sans-serif">Administrative Assistant</font></div><div><i><font face="arial, helvetica, sans-serif" color="#3d85c6"><b>Toyota Technological Institute</b></font></i></div><div><i><font face="arial, helvetica, sans-serif" color="#3d85c6">6045 S. Kenwood Avenue</font></i></div><div><i><font face="arial, helvetica, sans-serif" color="#3d85c6">Room 504</font></i></div><div><i><font face="arial, helvetica, sans-serif" color="#3d85c6">Chicago, IL 60637</font></i></div><div><i><font face="arial, helvetica, sans-serif">p:(773) 834-1757</font></i></div><div><i><font face="arial, helvetica, sans-serif">f: (773) 357-6970</font></i></div><div><b><i><a href="mailto:mmarre@ttic.edu" target="_blank"><font face="arial, helvetica, sans-serif">mmarre@ttic.edu</font></a></i></b></div></div></div></div></div></div></div></div></div></div>
<br><div class="gmail_quote">On Mon, Mar 13, 2017 at 2:24 PM, Mary Marre <span dir="ltr"><<a href="mailto:mmarre@ttic.edu" target="_blank">mmarre@ttic.edu</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div dir="ltr"><div style="font-size:12.8px"><div><div style="font-size:12.8px"><font color="#000000" face="arial, helvetica, sans-serif">When: Tuesday, March 14th at 11:00 am </font></div><div style="font-size:12.8px"><font color="#000000" face="arial, helvetica, sans-serif"><br></font></div><div style="font-size:12.8px"><font color="#000000" face="arial, helvetica, sans-serif">Where: <span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_8355097314499745974gmail-m_8816702620734345313gmail-m_3262824545120381495gmail-m_-1141671822915777344gmail-m_-7219251726624328345gmail-m_-8588148075564318222gmail-m_-8767966813928691312gmail-m_-1542318334608687154gmail-m_5717104778280916634gmail-m_4845490158781220632gmail-m_5124567205141626540gmail-m_3209361100497750746gmail-m_2953668934074478317gmail-m_-3155518689668024534m_9067904842688472155gmail-m_3071693547520408192gmail-il">TTIC</span>, 6045 S Kenwood Avenue, 5th Floor, Room 526</font></div><div style="font-size:12.8px"><font color="#000000" face="arial, helvetica, sans-serif"><br></font></div><div><span style="font-size:12.8px"><font color="#000000" face="arial, helvetica, sans-serif">Who: </font></span><font color="#000000" face="arial, helvetica, sans-serif">Nan Jiang, University of Michigan</font></div><div style="font-size:12.8px"><span style="font-size:12.8px"><font color="#000000" face="arial, helvetica, sans-serif"><br></font></span></div></div><span style="font-size:12.8px"></span></div><div style="font-size:12.8px"><br></div><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px">Title:<span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span></span><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">New</span></span><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> R</span></span><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">esults</span></span><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span></span><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">in</span></span><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> S</span></span><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px">tatistical</span><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> R</span></span><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px"><wbr>einforcement</span></span><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> L</span></span><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">earning</span></span><br class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><br class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px">Abstract:</span><br class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px">Recently,<span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span></span><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">reinforcement</span></span><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span></span><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px">learni<wbr>ng</span><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span>(RL) has achieved inspiring success<span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span></span><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">in</span></span><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span>game playing domains, including human-level control<span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span></span><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">in</span></span><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span>Atari games and mastering the game of Go. Looking into the future, we expect to build machine<span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span></span><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">learning</span></span><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span>systems that use RL to turn predictions into actions; applications include robotics, dialog systems, online education, adaptive medical treatment, to name but a few.<span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span></span><br class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><br class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">In</span></span><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span>this talk, I show how theoretical insights from supervised<span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span></span><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">learning</span></span><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span>can help understand RL, and better appreciate the unique challenges that arise from multi-stage decision making. The first part of the talk focuses on an interesting phenomenon, that a short planning horizon can produce better policies when there is limited data. I explain it by making a formal analogy to empirical risk minimization, and argue that a short planning horizon helps avoid overfitting. The second part of the talk concerns a core algorithmic challenge<span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span></span><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">in</span></span><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span>state-of-the-art RL: sample-efficient exploration<span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span></span><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">in</span></span><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span>large state spaces. I introduce a<span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span></span><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">new</span></span><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span>complexity measure, the Bellman rank, which allows us to apply a unified algorithm to a number of important RL settings,<span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span></span><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">in</span></span><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span>some cases obtaining polynomial sample complexity for the first time.</span><br class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><br class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px">Bio:</span><br class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-il">Nan</span> <span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-il">Jiang</span> is a PhD candidate<span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span></span><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">in</span></span><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span>Computer Science and Engineering at University of Michigan, He works with Satinder Singh on a variety of topics related to<span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span></span><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">reinforcement</span> </span><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"></span><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">learning</span></span><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px">. Specific research interests include provable use of function approximation, off-policy evaluation, state representation<span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span></span><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">learning</span></span><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px">, spectral<span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span></span><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">learning</span></span><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span>of dynamical systems, and inverse RL for AI safety. <span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-il">Nan</span> received his bachelor degree<span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span></span><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">in</span></span><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span>Control and Automation from Tsinghua University<span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span></span><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">in</span></span><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span>2011. He received the Best Paper Award at AAMAS 2015, and Rackham Predoctoral Fellowship<span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span></span><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">in</span></span><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604gmail-m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span>2016.</span><div style="font-size:12.8px"><font color="#212121" face="sans-serif"><br></font></div><div style="font-size:12.8px"><font color="#212121" face="sans-serif"><br></font></div><div style="font-size:12.8px"><font color="#212121" face="sans-serif">Host: <a href="mailto:mwalter@ttic.edu" target="_blank">Matthew Walter</a></font></div><div><br></div><div><br></div><div><br></div><div><br></div><div class="gmail_extra"><br clear="all"><div><div class="m_5023483748529811604gmail_signature" data-smartmail="gmail_signature"><div dir="ltr"><div><div dir="ltr"><div><div dir="ltr"><div><div dir="ltr"><font face="arial, helvetica, sans-serif">Mary C. Marre</font><div><font face="arial, helvetica, sans-serif">Administrative Assistant</font></div><div><i><font face="arial, helvetica, sans-serif" color="#3d85c6"><b>Toyota Technological Institute</b></font></i></div><div><i><font face="arial, helvetica, sans-serif" color="#3d85c6">6045 S. Kenwood Avenue</font></i></div><div><i><font face="arial, helvetica, sans-serif" color="#3d85c6">Room 504</font></i></div><div><i><font face="arial, helvetica, sans-serif" color="#3d85c6">Chicago, IL 60637</font></i></div><div><i><font face="arial, helvetica, sans-serif">p:<a href="tel:(773)%20834-1757" value="+17738341757" target="_blank">(773) 834-1757</a></font></i></div><div><i><font face="arial, helvetica, sans-serif">f: <a href="tel:(773)%20357-6970" value="+17733576970" target="_blank">(773) 357-6970</a></font></i></div><div><b><i><a href="mailto:mmarre@ttic.edu" target="_blank"><font face="arial, helvetica, sans-serif">mmarre@ttic.edu</font></a></i></b></div></div></div></div></div></div></div></div></div></div>
<br><div class="gmail_quote">On Wed, Mar 8, 2017 at 10:05 AM, Mary Marre <span dir="ltr"><<a href="mailto:mmarre@ttic.edu" target="_blank">mmarre@ttic.edu</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div dir="ltr"><div><div><div style="font-size:12.8px"><font color="#000000" face="arial, helvetica, sans-serif">When: Tuesday, March 14th at 11:00 am </font></div><div style="font-size:12.8px"><font color="#000000" face="arial, helvetica, sans-serif"><br></font></div><div style="font-size:12.8px"><font color="#000000" face="arial, helvetica, sans-serif">Where: <span class="m_5023483748529811604m_-5454150911817240326gmail-m_8355097314499745974gmail-m_8816702620734345313gmail-m_3262824545120381495gmail-m_-1141671822915777344gmail-m_-7219251726624328345gmail-m_-8588148075564318222gmail-m_-8767966813928691312gmail-m_-1542318334608687154gmail-m_5717104778280916634gmail-m_4845490158781220632gmail-m_5124567205141626540gmail-m_3209361100497750746gmail-m_2953668934074478317gmail-m_-3155518689668024534m_9067904842688472155gmail-m_3071693547520408192gmail-il">TTIC</span>, 6045 S Kenwood Avenue, 5th Floor, Room 526</font></div><div style="font-size:12.8px"><font color="#000000" face="arial, helvetica, sans-serif"><br></font></div><div><span style="font-size:12.8px"><font color="#000000" face="arial, helvetica, sans-serif">Who: </font></span><font color="#000000" face="arial, helvetica, sans-serif"><span style="font-size:12.8px">Nan Jiang, University of Michigan</span></font></div><div style="font-size:12.8px"><span style="font-size:12.8px"><font color="#000000" face="arial, helvetica, sans-serif"><br></font></span></div></div><span style="font-size:12.8px"></span></div><div><br></div><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px">Title:<span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span></span><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">New</span></span><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> R</span></span><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">esults</span></span><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span></span><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">in</span></span><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> S</span></span><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px">tatistical</span><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> R</span></span><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px"><wbr>einforcement</span></span><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> L</span></span><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">earning</span></span><br class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><br class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px">Abstract:</span><br class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px">Recently,<span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span></span><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">reinforcement</span></span><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span></span><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px">learni<wbr>ng</span><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span>(RL) has achieved inspiring success<span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span></span><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">in</span></span><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span>game playing domains, including human-level control<span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span></span><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">in</span></span><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span>Atari games and mastering the game of Go. Looking into the future, we expect to build machine<span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span></span><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">learning</span></span><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span>systems that use RL to turn predictions into actions; applications include robotics, dialog systems, online education, adaptive medical treatment, to name but a few.<span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span></span><br class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><br class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">In</span></span><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span>this talk, I show how theoretical insights from supervised<span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span></span><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">learning</span></span><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span>can help understand RL, and better appreciate the unique challenges that arise from multi-stage decision making. The first part of the talk focuses on an interesting phenomenon, that a short planning horizon can produce better policies when there is limited data. I explain it by making a formal analogy to empirical risk minimization, and argue that a short planning horizon helps avoid overfitting. The second part of the talk concerns a core algorithmic challenge<span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span></span><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">in</span></span><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span>state-of-the-art RL: sample-efficient exploration<span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span></span><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">in</span></span><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span>large state spaces. I introduce a<span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span></span><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">new</span></span><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span>complexity measure, the Bellman rank, which allows us to apply a unified algorithm to a number of important RL settings,<span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span></span><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">in</span></span><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span>some cases obtaining polynomial sample complexity for the first time.</span><br class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><br class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px">Bio</span><br class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="m_5023483748529811604m_-5454150911817240326gmail-il">Nan</span> <span class="m_5023483748529811604m_-5454150911817240326gmail-il">Jiang</span> is a PhD candidate<span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span></span><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">in</span></span><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span>Computer Science and Engineering at University of Michigan, He works with Satinder Singh on a variety of topics related to<span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span></span><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">reinforcement</span> </span><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"></span><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">learning</span></span><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px">. Specific research interests include provable use of function approximation, off-policy evaluation, state representation<span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span></span><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">learning</span></span><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px">, spectral<span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span></span><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">learning</span></span><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span>of dynamical systems, and inverse RL for AI safety. <span class="m_5023483748529811604m_-5454150911817240326gmail-il">Nan</span> received his bachelor degree<span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span></span><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">in</span></span><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span>Control and Automation from Tsinghua University<span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span></span><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">in</span></span><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span>2011. He received the Best Paper Award at AAMAS 2015, and Rackham Predoctoral Fellowship<span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span></span><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-lG m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px;background-color:rgba(251,246,167,0.498039);outline:transparent dashed 1px"><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-lG" style="outline:transparent dashed 1px">in</span></span><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg" style="color:rgb(33,33,33);font-family:sans-serif;font-size:13px"><span class="m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500inbox-inbox-m_1331964484770628044inbox-inbox-Apple-converted-space m_5023483748529811604m_-5454150911817240326gmail-m_7579992744826867500gmail_msg"> </span>2016.</span><div><font color="#212121" face="sans-serif"><br></font></div><div><font color="#212121" face="sans-serif"><br></font></div><div><font color="#212121" face="sans-serif">Host: <a href="mailto:mwalter@ttic.edu" target="_blank">Matthew Walter</a></font></div><div><br></div><div><font color="#212121" face="sans-serif"><br></font></div><div><font color="#212121" face="sans-serif"><br clear="all"></font><div><div class="m_5023483748529811604m_-5454150911817240326gmail_signature"><div dir="ltr"><div><div dir="ltr"><div><div dir="ltr"><div><div dir="ltr"><font face="arial, helvetica, sans-serif">Mary C. Marre</font><div><font face="arial, helvetica, sans-serif">Administrative Assistant</font></div><div><i><font face="arial, helvetica, sans-serif" color="#3d85c6"><b>Toyota Technological Institute</b></font></i></div><div><i><font face="arial, helvetica, sans-serif" color="#3d85c6">6045 S. Kenwood Avenue</font></i></div><div><i><font face="arial, helvetica, sans-serif" color="#3d85c6">Room 504</font></i></div><div><i><font face="arial, helvetica, sans-serif" color="#3d85c6">Chicago, IL 60637</font></i></div><div><i><font face="arial, helvetica, sans-serif">p:<a href="tel:(773)%20834-1757" value="+17738341757" target="_blank">(773) 834-1757</a></font></i></div><div><i><font face="arial, helvetica, sans-serif">f: <a href="tel:(773)%20357-6970" value="+17733576970" target="_blank">(773) 357-6970</a></font></i></div><div><b><i><a href="mailto:mmarre@ttic.edu" target="_blank"><font face="arial, helvetica, sans-serif">mmarre@ttic.edu</font></a></i></b></div></div></div></div></div></div></div></div></div></div>
</div></div>
</blockquote></div><br></div></div>
</blockquote></div><br></div></div>