Efficient algorithms for decision tree cross-validation

• DOI: 10.48550/arxiv.cs/0110036
• Publication Date: 2001-01-01
• Authors: Hendrik Blockeel; Jan Struyf
• Keywords: Cross-validation; Tree (set theory); Decision tree model

Cross-validation is a useful and generally applicable technique often employed in machine learning, including decision tree induction. An important disadvantage of straightforward implementation the its computational overhead. In this paper we show that, for trees, overhead cross-validation can be reduced significantly by integrating with normal induction process. We discuss how existing algorithms adapted to aim, provide an analysis speedups these adaptations may yield. The supported experimental results.