Rethinking Log Odds: Linear Probability Modelling and Expert Advice in Interpretable Machine Learning
This work addresses interpretability for users of machine learning models, particularly in financial domains, but it is incremental as it builds on existing additive models and expert advice algorithms.
The paper tackles the problem of interpretability in machine learning by introducing Linearised Additive Models (LAMs) and SubscaleHedge, which enhance interpretability without large performance penalties, as shown through rigorous testing on financial data with no significant drops in ROC-AUC and calibration.
We introduce a family of interpretable machine learning models, with two broad additions: Linearised Additive Models (LAMs) which replace the ubiquitous logistic link function in General Additive Models (GAMs); and SubscaleHedge, an expert advice algorithm for combining base models trained on subsets of features called subscales. LAMs can augment any additive binary classification model equipped with a sigmoid link function. Moreover, they afford direct global and local attributions of additive components to the model output in probability space. We argue that LAMs and SubscaleHedge improve the interpretability of their base algorithms. Using rigorous null-hypothesis significance testing on a broad suite of financial modelling data, we show that our algorithms do not suffer from large performance penalties in terms of ROC-AUC and calibration.