Cost-Sensitive Classification: Empirical Evaluation of a Hybrid Genetic Decision Tree Induction Algorithm

TL;DR

This paper presents ICET, a novel cost-sensitive classification algorithm using genetic algorithms to optimize decision tree biases, demonstrating superior performance on medical datasets compared to existing methods.

Contribution

Introduces ICET, a hybrid genetic decision tree induction algorithm that optimizes classification costs, outperforming existing cost-sensitive classifiers in empirical evaluations.

Findings

ICET outperforms EG2, CS-ID3, IDX, and C4.5 on medical datasets.

ICET maintains its advantage under various conditions.

Improved search method enhances ICET's bias optimization.

Abstract

This paper introduces ICET, a new algorithm for cost-sensitive classification. ICET uses a genetic algorithm to evolve a population of biases for a decision tree induction algorithm. The fitness function of the genetic algorithm is the average cost of classification when using the decision tree, including both the costs of tests (features, measurements) and the costs of classification errors. ICET is compared here with three other algorithms for cost-sensitive classification - EG2, CS-ID3, and IDX - and also with C4.5, which classifies without regard to cost. The five algorithms are evaluated empirically on five real-world medical datasets. Three sets of experiments are performed. The first set examines the baseline performance of the five algorithms on the five datasets and establishes that ICET performs significantly better than its competitors. The second set tests the robustness of ICET under a variety of conditions and shows that ICET maintains its advantage. The third set looks at ICET's search in bias space and discovers a way to improve the search.