Seminar: Kernel Sequential Monte Carlo

SpeakerIngmar Schuster
AffiliationUniversité Paris-Dauphine
DateFriday, 22 Apr 2016
Time13:00 - 14:00
LocationRoberts G08 Sir David Davies LT
Event seriesMicrosoft Research CSML Seminar Series

Bayesian posterior inference with Monte Carlo methods has a fundamental role in statistics and probabilistic machine learning. Target posterior distributions arising in increasingly complex models often exhibit high degrees of nonlinearity and multimodality and pose substantial challenges to traditional samplers.
We propose the Kernel Sequential Monte Carlo (KSMC) framework for building emulator models of the current particle system in a Reproducing Kernel Hilbert Space and use the emulator's geometry to inform local proposals. KSMC is applicable when gradients are unknown or prohibitively expensive and inherits the superior performance of SMC on multi-modal targets and its ability to estimate model evidence. Strengths of the proposed methodology are demonstrated on a series of challenging synthetic and real-world examples.

Joint work with Heiko Strathmann, Brooks Paige, Dino Sejdinovic.

Paper link

Speaker website

iCalendar csml_id_264.ics