SparkWeightedBlender

class sparklightautoml.automl.blend.SparkWeightedBlender(max_iters=5, max_inner_iters=7, max_nonzero_coef=0.05)[source]

Bases: SparkBlender, WeightedBlender

Weighted Blender based on coord descent, optimize task metric directly.

Weight sum eq. 1. Good blender for tabular data, even if some predictions are NaN (ex. timeout). Model with low weights will be pruned.