Inference Machine Learning; Probabilistic / Bayesian AI
noun
Definition: The process of using a trained model or a probabilistic program to compute outputs from inputs; in probabilistic and Bayesian contexts, inference specifically refers to computing posterior beliefs, latent-variable values, or query answers from a model and observed data [ Google ML Glossary].
Example in context: “Dealing with Continuous RVs Till now, we have not talked about inference in programs with continuous RVs.” [Kumar et al. 2023]
Synonyms: model inference; probabilistic inference (context-dependent)
Related terms: posterior inference; online inference; offline inference; prediction; reasoning