	CRIM	ZN	INDUS	CHAS	NOX	RM	AGE	DIS	RAD	TAX	PTRATIO	B	LSTAT	price
0	0.00632	18	2.31	0	0.538	6.575	65.2	4.09	1	296	15.3	396.9	4.98	24
1	0.02731	0	7.07	0	0.469	6.421	78.9	4.9671	2	242	17.8	396.9	9.14	21.6
2	0.02729	0	7.07	0	0.469	7.185	61.1	4.9671	2	242	17.8	392.83	4.03	34.7
3	0.03237	0	2.18	0	0.458	6.998	45.8	6.0622	3	222	18.7	394.63	2.94	33.4
4	0.06905	0	2.18	0	0.458	7.147	54.2	6.0622	3	222	18.7	396.9	5.33	36.2
5	0.02985	0	2.18	0	0.458	6.43	58.7	6.0622	3	222	18.7	394.12	5.21	28.7
6	0.08829	12.5	7.87	0	0.524	6.012	66.6	5.5605	5	311	15.2	395.6	12.43	22.9
7	0.14455	12.5	7.87	0	0.524	6.172	96.1	5.9505	5	311	15.2	396.9	19.15	27.1
8	0.21124	12.5	7.87	0	0.524	5.631	100	6.0821	5	311	15.2	386.63	29.93	16.5
9	0.17004	12.5	7.87	0	0.524	6.004	85.9	6.5921	5	311	15.2	386.71	17.1	18.9

2、数据预处理

2.1、分离特征和标签

data_X.shape,data_y.shape (506, 13) (506,)

2.2、获取类别型特征索引

categorical_features_index [3 8]

3、模型建立并训练

3.1、数据集切分

3.2、建立模型并训练

3.3、模型预测

RFR_R2： 0.7779596519110754
RFR_MSE： 2.445354901960794

4、解释模型

4.1、生成LIME解释器

Intercept 24.92879400416056
Prediction_local [22.50477631]
Right: 22.343499999999995

4.2、基于LIME对局部点的解释

测试集中任意挑选一个样本，设置用5个特征变量来开始解释。会输出线性代理模型的截距，系数，标准化之后的样本数据，用LIME得到的解释值以及RF预测该样本的值

exp_list
[('6.20 < RM <= 6.62', -2.795215449084299), ('7.19 < LSTAT <= 11.49', 1.844260653254622), ('19.05 < PTRATIO <= 20.20', -0.5871239693866448), ('330.00 < TAX <= 666.00', -0.5434185191566738), ('0.25 < CRIM <= 3.68', -0.34252040870538353)]

4.3、基于LIME显示详细信息图

解释的行以表格的形式显示在右侧，LIME在解释中离散化了特征

4.4、基于LIME显示权重图：输出对应的特征变量的局部解释图

输出对应的特征变量的局部解释图

4.5、利用SP-LIME带有子模块优化的LIME算法实现全局解释


………………-selector-tag">before -selector-tag">submodular_pick.SubmodularPick………………
-selector-tag">Intercept 26.57607003445879
-selector-tag">Prediction_local [17.28878941]
-selector-tag">Right: 13.99959999999987
-selector-tag">Intercept 24.46991901914869
-selector-tag">Prediction_local [22.71236873]
-selector-tag">Right: 41.66589999999996
-selector-tag">Intercept 25.23673616420291
-selector-tag">Prediction_local [21.2921698]
-selector-tag">Right: 21.184700000000095
-selector-tag">Intercept 26.426672242658093
-selector-tag">Prediction_local [18.86406084]
-selector-tag">Right: 18.158399999999947
-selector-tag">Intercept 25.57563422165384
-selector-tag">Prediction_local [22.20183543]
-selector-tag">Right: 24.299499999999707
-selector-tag">Intercept 24.14176081425899
-selector-tag">Prediction_local [25.52806146]
-selector-tag">Right: 22.002000000000127
-selector-tag">Intercept 19.958626465901155
-selector-tag">Prediction_local [37.24026643]
-selector-tag">Right: 46.02550000000028
-selector-tag">Intercept 24.71807610269299
-selector-tag">Prediction_local [23.61962036]
-selector-tag">Right: 31.206600000000346
-selector-tag">Intercept 25.600620511937866
-selector-tag">Prediction_local [20.21091994]
-selector-tag">Right: 17.617599999999797
-selector-tag">Intercept 24.523973775962826
-selector-tag">Prediction_local [22.87296153]
-selector-tag">Right: 19.997299999999775
………………-selector-tag">len(sp_obj.sp_explanations)……………… 1

网站声明：如果转载，请联系本站管理员。否则一切后果自行承担。

本文链接：https://www.xckfsq.com/news/show.html?id=1698

赞同 0

评论 0 条

ML之LIME：基于boston波士顿房价数据集回归预测利用LIME/SP-LIME局部解释图/权重图结合RF随机森林模型实现模型事后解释案例之详细攻略

基于boston波士顿房价数据集回归预测利用LIME/SP-LIME局部解释图/权重图结合RF随机森林模型实现模型事后解释案例之详细攻略

1、定义数据集