More from Events Calendar
- Mar 710:00 AMChemistry Student Seminar (CSS) - Xianjun "Cho" Zhang (Schlau-Cohen)Chemistry Student Seminar (CSS) is a student-organized seminar series that host graduate students and postdocs to share their research in a friendly and informal environment. Free donuts and coffee are provided.
- Mar 710:00 AMEnglish Conversation GroupAll sessions will take place on Zoom until March 7.On February 28, we’ll have a discussion "Black History Month."On March 7, we’ll be discussing "AI."Meet other MS&PC members from all over the world, get resources and information about life at MIT/Cambridge/Boston, exchange ideas, and engage in cultural conversations in a friendly and casual environment, while working on English fluency.Please email ecgatmit@gmail.com for more information.
- Mar 711:00 AMMen's Track and Field vs. New England DIII Indoor Track & Field ChampionshipTime: 11:00 AMLocation: Boston, MA / The TRACK at New Balance
- Mar 711:00 AMStochastics and Statistics SeminarSpeaker: Krishna Balasubramanian (University of California – Davis)Title: Finite-Particle Convergence Rates for Stein Variational Gradient DescentAbstract: Stein Variational Gradient Descent (SVGD) is a deterministic, interacting particle-based algorithm for nonparametric variational inference, yet its theoretical properties remain challenging to fully understand. This talk presents two complementary perspectives on SVGD. First, we introduce Gaussian-SVGD, a framework that projects SVGD onto the family of Gaussian distributions using a bilinear kernel. We establish rigorous convergence results for both mean-field dynamics and finite-particle systems, proving linear convergence to equilibrium in strongly log-concave settings. This framework also unifies recent algorithms for Gaussian Variational Inference (GVI) under a single theoretical lens. Second, we examine the finiteparticle convergence rates of nonparametric SVGD in Kernelized Stein Discrepancy (KSD) and Wasserstein-2 metrics. By decomposing the time derivative of relative entropy, we derive near-optimal convergence rates with polynomial dependence on dimensionality for certain kernel families. We also outline a framework to compare deterministic SVGD algorithms to the more standard randomized MCMC algorithms.Biography: Dr. Krishnakumar Balasubramanian works at the interface of statistics, optimization and machine learning with a focus on addressing inferential and computational aspects. He received his Ph.D from Georgia Tech. After appointments in Princeton University and and UW-Madison, he joined the Department of Statistics, UC Davis in Fall 2018. He has received a Facebook fellowship award and an ICML best paper runner-up award.
- Mar 711:00 AMWomen's Track and Field vs. New England DIII Indoor Track & Field ChampionshipTime: 11:00 AMLocation: Boston, MA / The TRACK at New Balance
- Mar 712:00 PMFilm Screening: Space is the PlaceAs part of the exhibition Afrofuturism and Otherworldliness, join us for a screening of the Afrofuturistic sci-fi film Space is the Place, directed by John Coney and featuring Sun Ra and his Arkestra. The film follows Sun Ra as he travels through time and space, seeking to transport Black people to a new utopian planet where they can thrive free from oppression. Blending surreal visuals, philosophical musings, and Sun Ra's signature music, this avant-garde masterpiece is part cosmic odyssey, part social commentary, and wholly unique. It's a kaleidoscopic exploration of liberation, identity, and the boundless power of imagination. Light snacks and popcorn will be provided.This event is presented as part of Artfinity: A celebration of creativity and community at MIT.