BEGIN:VCALENDAR VERSION:2.0 PRODID:-//wp-events-plugin.com//7.4.0.1//EN TZID:America/New_York X-WR-TIMEZONE:America/New_York BEGIN:VEVENT UID:897@fds.yale.edu DTSTART;TZID=America/New_York:20251103T153000 DTEND;TZID=America/New_York:20251103T170000 DTSTAMP:20251030T201202Z URL:https://fds.yale.edu/events/sds-seminar-claire-donnat-chicago/ SUMMARY:S&\;DS Seminar: Claire Donnat (Chicago)\, "Structured Topic Mod eling: Leveraging Sparsity and Graphs For Improved Inference" DESCRIPTION:\nAbstract: Classical topic models (LDA\, pLSI) treat documents as independent\, which wastes information when texts are short or vocabul aries are large. I will present two structured alternatives with statistic al guarantees. First\, a weakly sparse extension of pLSI that stabilizes e stimation in high-vocabulary settings by shrinking rare terms without enfo rcing hard zeros. Second\, a graph-aligned singular value decomposition th at incorporates known relationships between documents—e.g.\, spatial pro ximityor sample similarity—to improve recovery of document–topic and t opic–word matrices. For both methods we derive non-asymptotic\, high-pro bability error bounds for topic proportions and word distributions. Applic ations to spatial proteomics\, microbiome profiles\, and scientific abstra cts show accuracy and interpretability gains when side information is avai lable. The talk highlights when structure helps\, how to encode it\, and w hat guarantees are achievable.\n\n\n\nKeywords Topic Modeling · Constrain ed SVD\n\n\n\n\n CATEGORIES:FDS Events,Statistics & Data Science Seminar 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:STANDARD DTSTART:20251102T010000 TZOFFSETFROM:-0400 TZOFFSETTO:-0500 TZNAME:EST END:STANDARD END:VTIMEZONE END:VCALENDAR