编写用于创建决策树的程序有哪些步骤?

时间:2014-03-05 05:28:43

标签: python machine-learning cart-analysis

我想知道在Python中创建决策树(ID3)的步骤是什么?看起来如下的东西? (如果不是其他的话,或者它比那更好吗?)另外,为了创建像下面这样的ID3树,我应该计算熵?怎么样?

thal = fixed_defect [4 6]
|   ca <= 0.500000 [4 0]: negative
|   ca > 0.500000 [0 6]: positive
thal = normal [84 19]
|   thalach <= 111.500000 [0 4]: positive
|   thalach > 111.500000 [84 15]
|   |   age <= 55.500000 [56 4]
|   |   |   trestbps <= 113.500000 [9 3]
|   |   |   |   oldpeak <= 0.300000 [3 3]
|   |   |   |   |   cp = typ_angina [0 0]: negative
|   |   |   |   |   cp = asympt [0 2]: positive
|   |   |   |   |   cp = non_anginal [1 1]
|   |   |   |   |   |   age <= 44.000000 [1 0]: negative
|   |   |   |   |   |   age > 44.000000 [0 1]: positive
|   |   |   |   |   cp = atyp_angina [2 0]: negative
|   |   |   |   oldpeak > 0.300000 [6 0]: negative
|   |   |   trestbps > 113.500000 [47 1]
|   |   |   |   oldpeak <= 3.550000 [47 0]: negative
|   |   |   |   oldpeak > 3.550000 [0 1]: positive
|   |   age > 55.500000 [28 11]
|   |   |   chol <= 248.500000 [14 1]
|   |   |   |   oldpeak <= 2.800000 [14 0]: negative
|   |   |   |   oldpeak > 2.800000 [0 1]: positive
|   |   |   chol > 248.500000 [14 10]
|   |   |   |   sex = female [13 3]
|   |   |   |   |   cp = typ_angina [1 0]: negative
|   |   |   |   |   cp = asympt [3 3]
|   |   |   |   |   |   age <= 58.000000 [2 0]: negative
|   |   |   |   |   |   age > 58.000000 [1 3]
|   |   |   |   |   |   |   chol <= 362.000000 [0 3]: positive
|   |   |   |   |   |   |   chol > 362.000000 [1 0]: negative
|   |   |   |   |   cp = non_anginal [7 0]: negative
|   |   |   |   |   cp = atyp_angina [2 0]: negative
|   |   |   |   sex = male [1 7]
|   |   |   |   |   age <= 65.500000 [0 5]: positive
|   |   |   |   |   age > 65.500000 [1 2]
|   |   |   |   |   |   age <= 66.500000 [1 0]: negative
|   |   |   |   |   |   age > 66.500000 [0 2]: positive
thal = reversable_defect [20 67]
|   cp = typ_angina [3 1]
|   |   oldpeak <= 0.700000 [0 1]: positive
|   |   oldpeak > 0.700000 [3 0]: negative
|   cp = asympt [5 53]
|   |   oldpeak <= 0.650000 [5 10]
|   |   |   chol <= 240.500000 [5 2]
|   |   |   |   chol <= 192.000000 [1 2]
|   |   |   |   |   age <= 62.000000 [0 2]: positive
|   |   |   |   |   age > 62.000000 [1 0]: negative
|   |   |   |   chol > 192.000000 [4 0]: negative
|   |   |   chol > 240.500000 [0 8]: positive
|   |   oldpeak > 0.650000 [0 43]: positive
|   cp = non_anginal [9 10]
|   |   oldpeak <= 1.900000 [9 5]
|   |   |   trestbps <= 122.500000 [6 0]: negative
|   |   |   trestbps > 122.500000 [3 5]
|   |   |   |   chol <= 232.500000 [3 1]
|   |   |   |   |   trestbps <= 129.000000 [0 1]: positive
|   |   |   |   |   trestbps > 129.000000 [3 0]: negative
|   |   |   |   chol > 232.500000 [0 4]: positive
|   |   oldpeak > 1.900000 [0 5]: positive
|   cp = atyp_angina [3 3]
|   |   age <= 47.000000 [2 0]: negative
|   |   age > 47.000000 [1 3]
|   |   |   trestbps <= 109.000000 [1 0]: negative
|   |   |   trestbps > 109.000000 [0 3]: positive

0 个答案:

没有答案