acfx.abstract.ModelBasedCounterOptimizer

class acfx.abstract.ModelBasedCounterOptimizer

Bases: ABC

__init__()

abstract optimize_proba(target_class: int, feature_masked: List[str]) → Dict[str, float]

Modifies the instance to increase the probability of the target class by adjusting feature values.

Parameters

target_class:

The desired class to optimize towards.

feature_masked:

List of interchangeable features

Returns

Dictionary of feature names and their optimized values