[Colloquium] REMINDER: 11/6 TTIC Colloquium: Robert Nowak, University of Wisconsin-Madison

[Mary Marre via Colloquium]

When:     Monday, November 6th at 11:00 a.m.

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

Who:      Robert Nowak, University of Wisconsin-Madison



Title:      Outranked: Exploiting Nonlinear Algebraic Structure in Matrix
Recovery Problems

Abstract: This talk discusses two matrix recovery problems involving
nonlinear algebraic structure.  The first relates to learning low-rank
Euclidean embeddings and metrics from data, specifically from differences
or comparisons of distances.  The challenge here is that the (linear)
differencing operator has a nullspace.  We show that recovery is still
possible since, somewhat surprisingly, the differencing operator has a
nonlinear inverse when restricted to Euclidean distance matrices. The
second problem considers matrix completion in cases where columns lie in a
nonlinear algebraic variety (rather than the commonplace linear subspace
model). We propose a new algorithm for this problem based on data
tensorization in combination with standard low-rank matrix completion
methods. The challenge here is that the observation patterns in the
tensorized data representation are highly structured and far from uniformly
random.  We show that, under mild assumptions, the observation patterns are
generic enough to enable exact recovery.


Host: Karen Livescu <klivescu at ttic.edu>


For more information on the colloquium series or to subscribe to the
mailing list, please see http://www.ttic.edu/colloquium.php


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>*

On Sun, Nov 5, 2017 at 10:42 PM, Mary Marre <mmarre at ttic.edu> wrote:

> When:     Monday, November 6th at 11:00 a.m.
>
> Where:    TTIC, 6045 S. Kenwood Avenue, 5th Floor, Room 526
>
> Who:      Robert Nowak, University of Wisconsin-Madison
>
>
> Title:      Outranked: Exploiting Nonlinear Algebraic Structure in Matrix
> Recovery Problems
>
> Abstract: This talk discusses two matrix recovery problems involving
> nonlinear algebraic structure.  The first relates to learning low-rank
> Euclidean embeddings and metrics from data, specifically from differences
> or comparisons of distances.  The challenge here is that the (linear)
> differencing operator has a nullspace.  We show that recovery is still
> possible since, somewhat surprisingly, the differencing operator has a
> nonlinear inverse when restricted to Euclidean distance matrices. The
> second problem considers matrix completion in cases where columns lie in a
> nonlinear algebraic variety (rather than the commonplace linear subspace
> model). We propose a new algorithm for this problem based on data
> tensorization in combination with standard low-rank matrix completion
> methods. The challenge here is that the observation patterns in the
> tensorized data representation are highly structured and far from uniformly
> random.  We show that, under mild assumptions, the observation patterns are
> generic enough to enable exact recovery.
>
>
>
> Host: Karen Livescu <klivescu at ttic.edu>
>
>
> For more information on the colloquium series or to subscribe to the
> mailing list, please see http://www.ttic.edu/colloquium.php
>
>
> Mary C. Marre
> Administrative Assistant
> *Toyota Technological Institute*
> *6045 S. Kenwood Avenue*
> *Room 504*
> *Chicago, IL  60637*
> *p:(773) 834-1757 <(773)%20834-1757>*
> *f: (773) 357-6970 <(773)%20357-6970>*
> *mmarre at ttic.edu <mmarre at ttic.edu>*
>
> On Tue, Oct 31, 2017 at 10:22 AM, Mary Marre <mmarre at ttic.edu> wrote:
>
>> When:     Monday, November 6th at 11:00 a.m.
>>
>> Where:    TTIC, 6045 S. Kenwood Avenue, 5th Floor, Room 526
>>
>> Who:      Robert Nowak, University of Wisconsin-Madison
>>
>>
>>
>> Title:      Outranked: Exploiting Nonlinear Algebraic Structure in Matrix
>> Recovery Problems
>>
>> Abstract: This talk discusses two matrix recovery problems involving
>> nonlinear algebraic structure.  The first relates to learning low-rank
>> Euclidean embeddings and metrics from data, specifically from differences
>> or comparisons of distances.  The challenge here is that the (linear)
>> differencing operator has a nullspace.  We show that recovery is still
>> possible since, somewhat surprisingly, the differencing operator has a
>> nonlinear inverse when restricted to Euclidean distance matrices. The
>> second problem considers matrix completion in cases where columns lie in a
>> nonlinear algebraic variety (rather than the commonplace linear subspace
>> model). We propose a new algorithm for this problem based on data
>> tensorization in combination with standard low-rank matrix completion
>> methods. The challenge here is that the observation patterns in the
>> tensorized data representation are highly structured and far from uniformly
>> random.  We show that, under mild assumptions, the observation patterns are
>> generic enough to enable exact recovery.
>>
>>
>>
>> Host: Karen Livescu <klivescu at ttic.edu>
>>
>>
>> For more information on the colloquium series or to subscribe to the
>> mailing list, please see http://www.ttic.edu/colloquium.php
>>
>>
>> Mary C. Marre
>> Administrative Assistant
>> *Toyota Technological Institute*
>> *6045 S. Kenwood Avenue*
>> *Room 504*
>> *Chicago, IL  60637*
>> *p:(773) 834-1757 <(773)%20834-1757>*
>> *f: (773) 357-6970 <(773)%20357-6970>*
>> *mmarre at ttic.edu <mmarre at ttic.edu>*
>>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.cs.uchicago.edu/pipermail/colloquium/attachments/20171106/11d1fcea/attachment-0001.html>