experiments.experiment_utils.supervised_initial_solution_fn

supervised_initial_solution_fn(m, x, y)

A common initial solution function used in supervised learning. Just a wrapper for the model.fit() method.

Parameters:
  • m – SeldonianModel instance

  • x – features

  • y – labels

Returns:

The output of the fit() method, which are model weights as a flattened 1D array