Probabilistic Graphical Model (PGM) or Structured Probabilistic Model (SPM)
- is a type of probability distribution model that uses graphical models to systematically capture the behavior of a probability distribution
- is used to build models from data that are probabilistic/non-deterministic
- can be used for a wide range of tasks including simulated probabilities, making probabilistic inferences, prediction, anomaly detection, diagnostics, automated insight, reasoning, time series prediction, and decision-making under uncertainty.