<div dir="ltr">ICYMI I'll be proposing my thesis this Friday! Feel free to come by :)<div><br></div><div>Best,</div><div>Junier</div><div><br><div class="gmail_quote">---------- Forwarded message ----------<br>From: <b class="gmail_sendername">Diane Stidle</b> <span dir="ltr"><<a href="mailto:diane%2B@cs.cmu.edu">diane+@cs.cmu.edu</a>></span><br>Date: Fri, Mar 3, 2017 at 4:30 PM<br>Subject: Thesis Proposal - 3/10/17 - Junier Oliva - Distribution and Histogram (DisH) Learning<br>To: "<a href="mailto:ml-seminar@cs.cmu.edu">ml-seminar@cs.cmu.edu</a>" <<a href="mailto:ML-SEMINAR@cs.cmu.edu">ML-SEMINAR@cs.cmu.edu</a>>, Le Song <<a href="mailto:lsong@cc.gatech.edu">lsong@cc.gatech.edu</a>><br><br><br>
<div text="#000000" bgcolor="#FFFFFF">
<p><i>Thesis Proposal</i></p>
<p>Date: 3/10/17<br>
Time: 3:00pm<br>
Place: 8102 GHC<br>
Speaker: Junier Oliva</p>
<p>Title: Distribution and Histogram (DisH) Learning</p>
<p>Abstract:<br>
</p>
<div>
<div>This thesis advances the explicit use of distributions in
machine learning. We develop algorithms that consider
distributions as functional covariates/responses, and methods
that use distributions as internal representations. We consider
distributions since they are a straightforward characterization
of many natural phenomena and provide a richer description than
simple point data by detailing information at an aggregate
level. Our approach may be seen as addressing two sides of the
same coin: on one side, we use traditional machine learning
algorithms adjusted to directly operate on inputs and outputs
that are probability functions (and sample sets); on the other
side, we develop better estimators for traditional vector data
by making use of, and adjusting internal distributions.</div>
<div><br>
</div>
<div>We begin by developing algorithms for traditional machine
learning tasks for the cases when one's input (and/or possibly
output) is not a finite point, but is instead a distribution, or
sample set drawn from a distribution. We develop a scalable
nonparametric estimator for regressing a real valued response
given an input that is a distribution, a case which we coin
distribution to real regression (DRR). Furthermore, we extend
this work to the case when both the output response and the
input covariate are distributions; a task we call distribution
to distribution regression (DDR). Moreover, we propose flexible
and scalable techniques for conditional density estimation where
one regresses an output response that is a distribution given a
real valued covariate.</div>
<div><br>
</div>
<div>After, we look to expand the versatility and efficacy of
traditional machine learning tasks through novel methods that
operate with implicit or latent distributions. Take for example
kernel methods that use a shift-invariant kernel. Here, one's
kernel uniquely determines a distribution, the spectral density,
that controls the frequencies considered over inputs. We show
that one may improve the performance of kernel learning tasks by
learning this spectral distribution in a data-driven fashion
using Bayesian nonparametric techniques. Furthermore, we recast
classification as a task on distributions. Namely, the Bayes
classification risk is minimized when the distributions of
features of instances from each particular class are
non-overlapping. Hence, we propose a distribution based task
with ties to a Bayes risk to perform supervised feature
learning.</div>
<div><br>
</div>
<div>Leveraging the high-level, aggregate information provided by
distributions in these algorithms allows us to improve
performance in a broad range of domains including: cosmology,
neuroscience, computer vision, and natural language processing.
Furthermore, the scalable nature of our algorithms are such that
we may scale to millions, even billions of instances.<br>
<br>
Thesis Committee:<br>
Barnabas Poczos (Co-Chair)<br>
Jeff Schneider (Co-Chair)<br>
Ruslan Salakhutdinov<br>
Le Song (Georgia Institute of Technology)<br>
<br>
Link to draft document:<br>
<a class="m_-5944510183316843526moz-txt-link-freetext" href="https://www.dropbox.com/s/tbxj4v0oy35omky/Thesis_Proposal.pdf?dl=0" target="_blank">https://www.dropbox.com/s/<wbr>tbxj4v0oy35omky/Thesis_<wbr>Proposal.pdf?dl=0</a><span class="HOEnZb"><font color="#888888"><br>
</font></span></div><span class="HOEnZb"><font color="#888888">
</font></span></div><span class="HOEnZb"><font color="#888888">
<pre class="m_-5944510183316843526moz-signature" cols="72">--
Diane Stidle
Graduate Programs Manager
Machine Learning Department
Carnegie Mellon University
<a class="m_-5944510183316843526moz-txt-link-abbreviated" href="mailto:diane@cs.cmu.edu" target="_blank">diane@cs.cmu.edu</a>
<a href="tel:(412)%20268-1299" value="+14122681299" target="_blank">412-268-1299</a></pre>
</font></span></div>
</div><br></div></div>