Learning When Training Data are Costly: The Effect of Class Distribution on Tree Induction

TL;DR

This paper investigates how class distribution in training data affects the performance of classification trees, especially when training data is limited and costs are high, proposing a sampling algorithm to optimize class selection.

Contribution

It analyzes the impact of class distribution on tree induction performance and introduces a budget-sensitive sampling method for selecting training data to improve classifier accuracy.

Findings

Balanced class distribution improves ROC AUC performance.

Natural class distribution performs well with error rate evaluation.

Sampling algorithm yields classifiers with near-optimal performance.

Abstract

For large, real-world inductive learning problems, the number of training examples often must be limited due to the costs associated with procuring, preparing, and storing the training examples and/or the computational costs associated with learning from them. In such circumstances, one question of practical importance is: if only n training examples can be selected, in what proportion should the classes be represented? In this article we help to answer this question by analyzing, for a fixed training-set size, the relationship between the class distribution of the training data and the performance of classification trees induced from these data. We study twenty-six data sets and, for each, determine the best class distribution for learning. The naturally occurring class distribution is shown to generally perform well when classifier performance is evaluated using undifferentiated error rate (0/1 loss). However, when the area under the ROC curve is used to evaluate classifier performance, a balanced distribution is shown to perform well. Since neither of these choices for class distribution always generates the best-performing classifier, we introduce a budget-sensitive progressive sampling algorithm for selecting training examples based on the class associated with each example. An empirical analysis of this algorithm shows that the class distribution of the resulting training set yields classifiers with good (nearly-optimal) classification performance.