[KongchangAI]
Concept变分推断 / Variational Bayes

Variational Bayesian Methods

A class of Bayesian inference methods that approximate intractable posterior distributions with parameterized approximate distributions, trained by optimizing the evidence lower bound (ELBO). They support amortized, efficient inference.

Timeline (last 90 days)

Nov 3

Variational Bayesian methods approximate the posterior, which is difficult to compute exactly, with a parameterized approximate distribution, and are trained by optimizing the evidence lower bound (ELBO).

Unverified50%

All Facts (1)

Source Articles