[Colloquium] 12/12 Young Researcher Seminar Series: Junyoung Chung, Université de Montréal

Mary Marre via Colloquium colloquium at mailman.cs.uchicago.edu
Tue Dec 6 18:02:06 CST 2016


When:     Monday, November 12th at 11:00 am

Where:    TTIC, 6045 S. Kenwood Avenue, 5th Floor, Room 526

Who:       Junyoung Chung, Université de Montréal


Title:        Towards Hierarchical Multiscale Recurrent Neural Networks

Abstract: The recent resurgence of recurrent neural networks has led to
remarkable advances in various applications including machine translation,
speech recognition, speech synthesis and caption generation. However,
learning both hierarchical and temporal representation has been among the
longstanding challenges of recurrent neural networks. Multiscale recurrent
neural networks have been considered as a promising approach to resolve
this issue, yet there has been a lack of empirical evidence showing that
this type of models can actually capture the temporal dependencies by
discovering the latent hierarchical structure of the sequence.

In this talk, I will discuss about my previous works on multiscale
recurrent neural networks. I will show how a deep recurrent neural network
with extra gating units can update its layers with different timescales and
discover the underlying structure of the sequences. We will discuss about
three types of gated recurrent neural network architectures, gated feedback
recurrent neural networks that allow recurrent connection across different
level of recurrent layers via a soft gating mechanism, bi-scale recurrent
neural networks that consist with fast-changing and slow-changing recurrent
layers to model characters and words separately in neural machine
translation systems, and finally, hierarchical multiscale recurrent neural
networks that can capture the latent hierarchical structure of the
sequences with discrete gating units, so called boundary detectors. The
hierarchical multiscale recurrent neural networks can save notable amount
of computation compared to standard recurrent neural networks in principle
and can generalize better.

 host:* David McAllester <mcallester at ttic.edu>*


************************************************************
**************************************



The TTIC Young Researcher Seminar Series (http://www.ttic.edu/young-
researcher.php) features talks by Ph.D. students and postdocs whose research is
of broad interest to the computer science community. The series provides an
opportunity for early-career researchers to present recent work to and meet
with students and faculty at TTIC and nearby universities.


The seminars are typically held on Wednesdays at 11:00am in TTIC Room 526.

For additional information, please contact Matthew Walter (mwalter at ttic.edu
).




Mary C. Marre
Administrative Assistant
*Toyota Technological Institute*
*6045 S. Kenwood Avenue*
*Room 504*
*Chicago, IL  60637*
*p:(773) 834-1757*
*f: (773) 357-6970*
*mmarre at ttic.edu <mmarre at ttic.edu>*
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.cs.uchicago.edu/pipermail/colloquium/attachments/20161206/8f56e79b/attachment.html>


More information about the Colloquium mailing list