BEGIN:VCALENDAR
VERSION:2.0
PRODID:icalendar-ruby
CALSCALE:GREGORIAN
X-WR-CALNAME:Dynamics in Algorithm Design: Optimization\, Sampling and Diff
 usion
X-WR-TIMEZONE:Eastern Time (US & Canada)
BEGIN:VEVENT
DTSTAMP:20260812T080639Z
UID:tag:localist.com\,2008:EventInstance_45925371121929
DTSTART:20240321T130000Z
DTEND:20240321T140000Z
DESCRIPTION:Optimization plays an important role across data science\, from
  guiding decision-making under uncertainty\, to modeling equilibrium in ga
 me dynamics. In this talk\, I will begin by describing a line of work stud
 ying information-theoretic complexity in optimization using higher-order d
 erivatives and parallel access to the gradient of the objective function\,
  aiming to shed light on efficient algorithm design under these novel comp
 utational models. \n\nSampling\, on the other hand\, is a fundamental task
  underlying many scientific pursuits\, from molecular dynamics to Lattice 
 QCD simulations. In the second part of the talk\, I will introduce ways to
  conceptualize designing dynamics in infinite-dimensional metric spaces\, 
 leading to geometry-aware MCMC sampling algorithms and controlled diffusio
 n processes that adopt a learning-driven approach for such numerical tasks
 .   \n\nI will end by highlighting (1) theoretically\, the connection betw
 een optimization\, sampling\, physics-inspired dynamical system\, mean-fie
 ld game goes much deeper than one may expect\; (2) computationally\, bring
 ing powerful function fitting NN-architecture to solve more traditional ta
 sks in PDE\, sampling\, control etc.\, offers many exciting opportunities 
 for both methodology developments and applications.\n\nPierce 301 or Zoom 
 (password:  687666)
GEO:42.3783;-71.117081
LOCATION:Pierce Hall\, Pierce 301
SUMMARY:Dynamics in Algorithm Design: Optimization\, Sampling and Diffusion
URL;VALUE=URI:https://events.seas.harvard.edu/event/dynamics-in-algorithm-d
 esign-optimization-sampling-and-diffusion
CATEGORIES:Colloquia / Seminar / Lecture
END:VEVENT
END:VCALENDAR