<div dir="ltr">A big day coming up for the Lab, please mark your calendars and come listen to an exciting talk by Cristian!<div><br></div><div>Cheers</div><div>Artur<br><br><div class="gmail_quote"><div dir="ltr" class="gmail_attr">---------- Forwarded message ---------<br>From: <strong class="gmail_sendername" dir="auto">Diane Stidle</strong> <span dir="auto"><<a href="mailto:stidle@andrew.cmu.edu">stidle@andrew.cmu.edu</a>></span><br>Date: Mon, Mar 18, 2024 at 3:29 PM<br>Subject: Thesis Defense - April 2, 2024 - Cristian Challu - Representing Time: Towards Pragmatic Multivariate Time Series Modeling<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>>, <<a href="mailto:ywu@stat.ucla.edu">ywu@stat.ucla.edu</a>><br></div><br><br><u></u>
<div>
<p><b><i>Thesis Defense</i></b></p>
<p>Date: April 2, 2024<br>
Time: 11:00am (EDT)<br>
Place: NSH 3305 & Remote<br>
PhD Candidate: Cristian Challu</p>
<p><b>Title: Representing Time: Towards Pragmatic Multivariate Time
Series Modeling</b></p>
<p>Abstract:<br>
Time series models are specialized in learning temporal
dependencies among observations and interactions between features
in a data stream. During the last decade, the unprecedented
success of deep learning models on Computer Vision and Natural
Language Processing has steadily permeated to time series tasks.
From Recurrent Neural Networks to Transformers, new advancements
in architectural design improved capabilities and performance.
Despite this success, I identify several challenges to adopting
current state-of-the-art methods, including handling distribution
shifts and missing data, computational complexity, and
interpretability.</p>
<div>The success of DL models is usually explained by their ability
to automatically discover helpful data representations.
Multivariate time series models involve high-dimensional objects
with numerous time series and temporal observations. However, they
often exhibit strong temporal dependencies and inter-feature
relations. In this thesis, I propose to design algorithms for
forecasting and anomaly detection tasks that leverage these
dependencies to induce efficient learning of representations that
satisfy desirable properties that can (i) improve the models'
performance, (ii) improve robustness by favoring domain
adaptation, and (iii) reduce overparameterization to improve
scalability. The completed work includes three parts, presenting
seven models and algorithms that achieve SoTA performance in
various tasks while addressing key adoption challenges.<br>
<br>
In the first part, I explore the dynamic latent space principle
and design latent temporal representations to make robust anomaly
detection and forecasting models. In the second part, I present a
novel scalable and interpretable model for multi-step forecasting
based on a non-linear frequency decomposition with connections to
Wavelet theory. It also features two extensions on using
multivariate exogenous covariates for high-impact domains,
including energy and healthcare. Finally, in the third part, I
present a large-scale study on enabling conditions, on both model
design and data characteristics, for transferability of
pre-trained models on time series tasks.<br>
</div>
<div><br>
<div><b>Thesis Committee:</b></div>
<div>Artur Dubrawski, Chair</div>
<div>Roni Rosenfeld</div>
<div>Barnabas Poczos</div>
Ying Nian Wu (UCLA)</div>
<div><br>
</div>
<div>Link to Draft Document: <a href="https://drive.google.com/file/d/17luIWYw3bLRYe-BuXn1QstlQad0ozXu8/view?usp=sharing" target="_blank">https://drive.google.com/file/d/17luIWYw3bLRYe-BuXn1QstlQad0ozXu8/view?usp=sharing</a></div>
<div><br>
</div>
<div>Zoom meeting link: <br>
</div>
<div><a href="https://cmu.zoom.us/j/99919622332?pwd=Y0E0ZU0xMCswbmFyTXBzNDRuK05mZz09" target="_blank">https://cmu.zoom.us/j/99919622332?pwd=Y0E0ZU0xMCswbmFyTXBzNDRuK05mZz09</a><br>
<br>
Meeting ID: 999 1962 2332<br>
Passcode: 902254<br>
</div>
<div><br>
</div>
<p></p>
<pre cols="72">--
Diane Stidle
PhD Program Manager
Machine Learning Department
Carnegie Mellon University
<a href="mailto:stidle@andrew.cmu.edu" target="_blank">stidle@andrew.cmu.edu</a></pre>
</div>
</div></div></div>