<div dir="auto">Team, please mark your calendars for Mononito's upcoming thesis proposal talk.<div dir="auto"><br></div><div dir="auto">Cheers</div><div dir="auto">Artur</div></div><br><div class="gmail_quote"><div dir="ltr" class="gmail_attr">---------- Forwarded message ---------<br>From: <strong class="gmail_sendername" dir="auto">Suzanne Muth</strong> <span dir="auto"><<a href="mailto:lyonsmuth@cmu.edu">lyonsmuth@cmu.edu</a>></span><br>Date: Mon, Apr 22, 2024, 9:47 AM<br>Subject: RI Ph.D. Thesis Proposal: Mononito Goswami<br>To: RI People <<a href="mailto:ri-people@andrew.cmu.edu">ri-people@andrew.cmu.edu</a>><br></div><br><br><div dir="ltr"><div class="gmail_default"><span style="color:rgb(80,0,80)"><div><font face="arial, sans-serif"><p style="margin:0in;color:rgb(0,0,0)">Date: 01 May 2024<br>Time: 2:00 PM (ET)<br>Location: <span style="color:rgb(34,34,34)">NSH 1305</span></p><p style="margin:0in;color:rgb(0,0,0)">Zoom Link: <span style="color:rgb(34,34,34)"><a href="https://cmu.zoom.us/j/97119194168?pwd=Uzd4cXFzTHRuZWU1RVJwb1JOMHU0UT09" target="_blank" rel="noreferrer">https://cmu.zoom.us/j/97119194168?pwd=Uzd4cXFzTHRuZWU1RVJwb1JOMHU0UT09</a></span></p></font></div></span><div><span style="color:rgb(80,0,80)"><font face="arial, sans-serif"><p style="margin:0in;color:rgb(0,0,0)">Type: Ph.D. Thesis Proposal<br>Who: Mononito Goswami<br>Title: <span style="color:rgb(34,34,34)">Towards Pragmatic Time Series Intelligence</span><br></p><p style="margin:0in;color:rgb(0,0,0)"><br></p><p style="margin:0in;color:rgb(0,0,0)">Abstract:</p></font></span><p style="margin:0in;color:rgb(0,0,0)"><font face="arial, sans-serif">The widespread adoption of time series machine learning (ML) models faces multiple challenges involving data, modeling and evaluation.<br><br><i>Data.</i> Modern ML models depend on copious amounts of cohesive and reliably annotated data for training and evaluation. However, labeled data is not always available and reliable, and can also be dispersed across different locations. We propose systematic solutions to making time series data ML-ready.<br><br><i>Modeling.</i> Most current time series ML models are built, trained and evaluated on individual datasets from a specific application domain. Thus, to build an effective model for a particular application scenario, substantial effort, time, and domain expertise are required to develop a successful task-specific design. We propose to partially address this limitation by developing large pre-trained foundation models for time series, to ease development of useful models across diverse application domains with limited resources, data and labels.<br><br><i>Evaluation.</i> Currently, time series models are commonly evaluated using relatively small, specific and highly tailored benchmarks, which may obfuscate assessment of their performance. We highlight the gaps in evaluation techniques and propose addressing the most important of them through comprehensive, multi-metric assessment.<br><br>In summary, this thesis aims to democratize time series artificial intelligence by simplifying and accelerating development of models, while improving their performance in real-world application scenarios facing resource constraints and imperfect data.<br><br>Thesis Committee Members:<br>Artur Dubrawski, Chair<br>Jean Oh<br>Barnabas Poczós<br>Frederic Sala, University of Wisconsin-Madison<br>Laurent Callot, Amazon<br></font></p><p style="margin:0in;color:rgb(0,0,0)"><font face="arial, sans-serif"><br></font></p><p style="margin:0in;color:rgb(0,0,0)"><font face="arial, sans-serif">A draft of the thesis proposal document is available at: </font></p><p style="margin:0in;color:rgb(0,0,0)"><a href="https://drive.google.com/drive/folders/1yzdjxg0Koj92msch3kRxJDemi29E9Snj?usp=sharing" target="_blank" rel="noreferrer"><font face="arial, sans-serif">https://drive.google.com/drive/folders/1yzdjxg0Koj92msch3kRxJDemi29E9Snj?usp=sharing</font></a></p></div></div></div>
</div>