Generative model

A type of statistical model where the researcher estimates a ‘general’ joint probability distribution

$P(X,Y)$ Where $X$ is the inputs (or hidden states) and $Y$ is the outputs or labels.

Once this joint probability function is determined, Bayes theorem can be used $P(Y|X)=\frac{P(X,Y)}{P(X)}=\frac{P(X|Y)P(Y)}{P(X)}$

Then in order to make the prediction the chose the label Y with the highest conditional probability

---

References

@ng2001 Contrast to Discriminative model