Algorithms
Methods for training a parametric model.
The algorithms aims to solve a stochastic minimization problem in order to recover the optimal model:
\[\min_\theta E_{(x,y) \sim P} [f(x,y,\theta)] \approx \min_\theta \sum_{i=1}^N f(x_i,y, \theta)\]
Note that \(y\) is optional.