Lecture

Belief Propagation on Graphs

In course
DEMO: culpa veniam fugiat
Elit non laborum irure eu enim sit culpa minim Lorem. Incididunt sit veniam tempor sit ullamco Lorem sit nulla in ullamco cupidatat id. Consequat duis elit sit exercitation mollit id anim laboris occaecat amet in culpa qui do. Officia qui exercitation quis exercitation Lorem do mollit elit id consequat minim. Culpa reprehenderit incididunt adipisicing consectetur id. Amet irure nulla ea ex ut adipisicing anim irure excepteur consequat.
Login to see this section
Description

This lecture introduces the concept of belief propagation on graphs, focusing on computing the partition function and free energy of statistical physics models. The instructor explains how to iterate the belief propagation equations on trees and extends the method to loopy graphs, discussing the challenges and heuristics involved. The lecture delves into the computation of free entropy and the implications of the independence assumption in incoming messages. It also explores the application of belief propagation on sparse random graphs, highlighting the significance of long loops in such graphs. The instructor demonstrates how the expected length of loops in sparse random graphs grows logarithmically with the graph size, providing insights into the feasibility and limitations of belief propagation algorithms.

This video is available exclusively on Mediaspace for a restricted audience. Please log in to MediaSpace to access it if you have the necessary permissions.

Watch on Mediaspace
Instructors (2)
officia et
Incididunt fugiat aliquip nostrud cillum exercitation et dolore amet consequat. Veniam commodo in culpa pariatur enim adipisicing nulla consequat qui consequat nisi. Sint tempor cillum sunt nostrud. Lorem commodo sunt exercitation irure incididunt ipsum nulla Lorem. Pariatur amet dolor Lorem in reprehenderit velit elit consequat.
dolore officia deserunt
Ut id nulla ut voluptate eiusmod tempor aliqua eu adipisicing elit sunt voluptate aliquip fugiat. Incididunt velit cupidatat exercitation minim irure excepteur id voluptate et. Eu dolor irure eiusmod dolor. Culpa pariatur non minim officia id cupidatat adipisicing tempor. Excepteur aliqua et nisi sint tempor occaecat tempor. Ea adipisicing sit eiusmod ipsum adipisicing nulla id sunt. Exercitation laboris eiusmod dolore mollit nostrud amet sint nulla.
Login to see this section
About this result
This page is automatically generated and may contain information that is not correct, complete, up-to-date, or relevant to your search query. The same applies to every other page on this website. Please make sure to verify the information with EPFL's official sources.
Related lectures (37)
Analyzing Limits: Indeterminate Forms
Explores handling indeterminate forms in limits through simplification and extracting dominant terms for effective evaluation.
Learning from the Interconnected World with Graphs
Explores learning from interconnected data using graphs, covering challenges, GNN design, research landscapes, and democratization of Graph ML.
Graph Neural Networks: Interconnected World
Explores learning from interconnected data with graphs, covering modern ML research goals, pioneering methods, interdisciplinary applications, and democratization of graph ML.
Understanding Surface Integrals
Explores surface integrals, emphasizing physical interpretation and mathematical calculations in vector fields and domains.
Automorphism Groups: Essential Chief Series
Explores essential chief series in tdlc groups, focusing on closed, normal subgroups and their chief factors.
Show more

Graph Chatbot

Chat with Graph Search

Ask any question about EPFL courses, lectures, exercises, research, news, etc. or try the example questions below.

DISCLAIMER: The Graph Chatbot is not programmed to provide explicit or categorical answers to your questions. Rather, it transforms your questions into API requests that are distributed across the various IT services officially administered by EPFL. Its purpose is solely to collect and recommend relevant references to content that you can explore to help you answer your questions.