决策树算法是数据挖掘中非常活跃的研究领域。通过对数据挖掘中决策树的基本思想进行阐述,讨论了决策树经典算法(ID3算法)的计算复杂度问题,并针对这一问题提出了利用统计理论知识和条件概率的思想来改进构造决策树的算法。实验表明,这种构造决策树算法的计算复杂度明显优于传统的算法,其效率也有很大的提高。 Decision tree algorithm is a very active research field in data mining. This paper describes the basic idea of ​​decision tree in data mining and discusses the computational complexity of the decision tree classic algorithm (ID3 algorithm). In view of this problem, the idea of ​​using statistical theory knowledge and conditional probability is proposed to improve the construction decision tree Algorithm. Experiments show that the computational complexity of this construction decision tree algorithm is obviously better than the traditional algorithm, and its efficiency is also greatly improved.