+A **Discriminative Model** directly learns the boundary that separates different classes of data, rather than modeling the distribution of each class. It focuses on predicting an output label `y` given input features `x`, often optimizing for tasks like [Classification](/wiki/classification) or [Regression](/wiki/regression). This approach directly models the conditional probability $P(y|x)$, contrasting with [Generative Model](/wiki/generative_model)s.
+## See also
+- [Generative Model](/wiki/generative_model)
+- [Machine Learning](/wiki/machine_learning)
+- [Classification](/wiki/classification)
... 1 more lines