Why is the prediction wrong? Towards underfitting case explanation via meta-classification

TL;DR

This paper introduces a heuristic method that explains misclassified data points by projecting them into interpretable meta-representations and using a meta-classifier to diagnose whether errors are due to model underfitting or data inseparability.

Contribution

The work presents a novel approach combining meta-representations and meta-classification to interpret individual misclassifications, especially underfitting cases, with high accuracy and transferability.

Findings

Achieved over 80% diagnosis label accuracy.

Meta-representation provides invariance across classifiers and datasets.

Meta-classifier effectively distinguishes between underfitting and data overlap errors.

Abstract

In this paper we present a heuristic method to provide individual explanations for those elements in a dataset (data points) which are wrongly predicted by a given classifier. Since the general case is too difficult, in the present work we focus on faulty data from an underfitted model. First, we project the faulty data into a hand-crafted, and thus human readable, intermediate representation (meta-representation, profile vectors), with the aim of separating the two main causes of miss-classification: the classifier is not strong enough, or the data point belongs to an area of the input space where classes are not separable. Second, in the space of these profile vectors, we present a method to fit a meta-classifier (decision tree) and express its output as a set of interpretable (human readable) explanation rules, which leads to several target diagnosis labels: data point is either correctly classified, or faulty due to a too weak model, or faulty due to mixed (overlapped) classes in the input space. Experimental results on several real datasets show more than 80% diagnosis label accuracy and confirm that the proposed intermediate representation allows to achieve a high degree of invariance with respect to the classifier used in the input space and to the dataset being classified, i.e. we can learn the metaclassifier on a dataset with a given classifier and successfully predict diagnosis labels for a different dataset or classifier (or both).