增量决策树算法及复杂度分析

An Incremental Alogrithm for Inducing Decision Trees and Its Complexity

摘要: 介绍了增量决策树算法的基本原理,并从实例费用和信息熵费用两个角度出发,对增量决策树算法的复杂度进行分析。通过实例说明,增量决策树算法能够构造出与ID3算法形态基本相同的决策树。

Abstract: An incremental algorithm for inducing decision trees is presented based on ID3 algorithm. The complexity of the incremental algorithm is analyzed in terms of instance-count additions and e-score calculations. The same training instance shows that the incremental algorithm can induce decision trees equivalent to those forms by ID3 algorithm.