BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//wp-events-plugin.com//7.2.3.1//EN
TZID:America/New_York
X-WR-TIMEZONE:America/New_York
BEGIN:VEVENT
UID:840@fds.yale.edu
DTSTART;TZID=America/New_York:20250409T113000
DTEND;TZID=America/New_York:20250409T130000
DTSTAMP:20250916T142150Z
URL:https://fds.yale.edu/events/fds-colloquium-munther-dahleh-mit/
SUMMARY:FDS Colloquium: Munther Dahleh (MIT)\, "Learning multiple models wi
 th limited and sparse data"
DESCRIPTION:\nAbstract: In this talk\, we address the challenge of learning
  models for a large number of agents when only limited data is available f
 or each individual.  We assume the existence of an underlying shared stru
 cture across agents\, enabling us to leverage their commonalities to impro
 ve model estimation. This setting arises in various applications\, includi
 ng healthcare\, economics\, and finance.\n\n\n\nWe propose two shared-stru
 cture modeling approaches. The first assumes that all agent models lie wit
 hin a low-rank subspace. The second extends this idea to the case of mixtu
 re models\, where each agent is modeled as a random sample of a fixed numb
 er of models. A central challenge in this setting is that the lack of suff
 icient data per agent makes it difficult to directly estimate their models
  or identify clusters of similar agents.\n\n\n\nTo address this\, we intro
 duce a framework that harnesses data across all agents to learn the shared
  structure—and in some cases\, the individual agent models. Building on 
 spectral methods for mixtures of linear regressions\, we present a moment-
 based estimator that recovers the underlying low-rank subspace. For mixtur
 e models\, we further employ tensor decomposition to estimate the mixture 
 components.\n\n\n\nOur approach generalizes to more complex settings\, inc
 luding mixtures of autoregressive (AR) processes and input-output (I/O) sy
 stems. We also demonstrate how both deterministic and stochastic dynamics 
 can be handled within our framework.\n\n\n\nThis work is in collaboration 
 with Maryann Rui.\n\n\n\nSpeaker bio: Munther Dahleh is the William A. Coo
 lidge Professor in Electrical Engineering and Computer Science (EECS) and 
 a member MIT’s Laboratory for Information and Decision Systems (LIDS). D
 ahleh was the founding director of the Institute for Data\, Systems\, and 
 Society (IDSS)\, serving from July 1\, 2015 to June 30\, 2023. He was prev
 iously the associate department head of EECS.\n\n\n\nProf. Dahleh joined L
 IDS as an assistant professor of EECS in 1987 and became a full professor 
 in 1998. He spent the spring of 1993 as a visiting professor in the Depart
 ment of Electrical Engineering\, California Institute of Technology\, and 
 has held consulting positions with several companies in the U.S. and abroa
 d.\n\n\n\nProf. Dahleh is internationally known for his fundamental contri
 butions to robust control theory\, computational methods for controller de
 sign\, the interplay between information and control\, the fundamental lim
 its of learning and decision in networked systems\, and the detection and 
 mitigation of systemic risk in interconnected and networked systems.\n\n\n
 \nIn particular\, his research interests include:\n\n\n\n-Transportation S
 ystems: Dynamic models of congestion under disruptions\, dependence of fra
 gility on network topology\, cascaded failures\, value of side information
  on performance\n\n\n\n-Networked Systems: Foundational theory for the int
 eraction between physical and information networks\, Information propagati
 on\, distributed decisions\, learning network structure from data.\n\n\n\n
 -Social Networks: Information cascades in stochastic networks\, opinion dy
 namics\, global games in modeling outcomes of crises\n\n\n\n-Systemic Risk
 : The development of a foundational theory for the early detection and con
 trol of systemic risk resulting from idiosyncratic disturbance affecting c
 omponents of a networked system\n
CATEGORIES:FDS Events,Colloquium
LOCATION:Yale Institute for Foundations of Data Science\, Kline Tower 13th 
 Floor\, Room 1327\, New Haven\, CT\, 06511\, United States
X-APPLE-STRUCTURED-LOCATION;VALUE=URI;X-ADDRESS=Kline Tower 13th Floor\, Ro
 om 1327\, New Haven\, CT\, 06511\, United States;X-APPLE-RADIUS=100;X-TITL
 E=Yale Institute for Foundations of Data Science:geo:0,0
END:VEVENT
BEGIN:VTIMEZONE
TZID:America/New_York
X-LIC-LOCATION:America/New_York
BEGIN:DAYLIGHT
DTSTART:20250309T030000
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
TZNAME:EDT
END:DAYLIGHT
END:VTIMEZONE
END:VCALENDAR