Normalizing Constants: Turning Neural Network Scores into Probability Distributions

Automated excerpt

It becomes a valid probability density only after normalization. Approximate the Normalizing Constant: Energy-Based Models approach the difficult Normalizing Constant through approximation. For flexible neural-network probability models, computing it over the entire input space may become intractable.

Selected automatically from source text; not independently written or fact-checked. Read the original for full context.