学习资料：

1. https://blog.csdn.net/choven_meng/article/details/82878018?ops_request_misc=%257B%2522request%255Fid%2522%253A%2522159566999819195264530507%2522%252C%2522scm%2522%253A%252220140713.130102334.pc%255Fall.%2522%257D&request_id=159566999819195264530507&biz_id=0&utm_medium=distribute.pc_search_result.none-task-blog-2_allfirst_rank_ecpm_v3~pc_rank_v2-2-82878018.first_rank_ecpm_v3_pc_rank_v2&utm_term=id3%E5%86%B3%E7%AD%96%E6%A0%91%E5%92%8Ccart%E5%86%B3%E7%AD%96%E6%A0%91&spm=1018.2118.3001.4187

2. https://blog.csdn.net/u010089444/article/details/53241218?ops_request_misc=%257B%2522request%255Fid%2522%253A%2522159566999919725247610063%2522%252C%2522scm%2522%253A%252220140713.130102334…%2522%257D&request_id=159566999919725247610063&biz_id=0&utm_medium=distribute.pc_search_result.none-task-blog-2_allbaidu_landing_v2~default-1-53241218.first_rank_ecpm_v3_pc_rank_v2&utm_term=id3%E5%86%B3%E7%AD%96%E6%A0%91%E5%92%8Ccart%E5%86%B3%E7%AD%96%E6%A0%91&spm=1018.2118.3001.4187

3. https://blog.csdn.net/qq_27782503/article/details/89064624

关键点：

1. 经验熵、经验条件熵、基尼系数的计算方法，由此推出构建决策树的逻辑。
2. ID3决策树只能处理离散型数据且用于分类；C4.5可以处理离散型也可以处理连续型数据，且用于分类；CART决策树可以做回归，也可以做分类，连续型数据和离散型数据均可处理。