机器学习笔记

1.ID3决策树算法

https://blog.youkuaiyun.com/qq_36330643/article/details/77415451

2.C4.5决策树算法

https://blog.youkuaiyun.com/zhihua_oba/article/details/70632622

3.GBDT详解

基于残差的gbdt是一种特殊的gbdt模型，它的损失函数是平方损失函数，常用来处理回归类的问题。

https://zhuanlan.zhihu.com/p/29765582

https://www.cnblogs.com/peizhe123/p/5086128.html

4.CART决策树算法

https://www.cnblogs.com/yonghao/p/5135386.html

5.朴素贝叶斯算法

https://blog.youkuaiyun.com/shengweisong/article/details/52750907

6.logistic回归

https://www.cnblogs.com/home123/p/7356523.html

7.异常检测（欺诈）

http://baijiahao.baidu.com/s?id=1599068399235692908&wfr=spider&for=pc

8.数据预处理

https://www.cnblogs.com/dshn/p/8856173.html

9.凝聚层次聚类

https://blog.youkuaiyun.com/u012500237/article/details/65437525

10,DBSCAN算法

https://blog.youkuaiyun.com/huacha__/article/details/81094891

11.classification_report&精确度/召回率/F1值

https://blog.youkuaiyun.com/akadiao/article/details/78788864

12.xgboost进行二分类或者多分类

https://blog.youkuaiyun.com/kwame211/article/details/81098025

https://blog.youkuaiyun.com/wang1127248268/article/details/76769016

https://blog.youkuaiyun.com/kwame211/article/details/81098025

13.稀疏特征

https://www.cnblogs.com/yifdu25/p/8128028.html

https://blog.youkuaiyun.com/diligent_321/article/details/88138393

14.正则化

https://blog.youkuaiyun.com/speargod/article/details/80233619

15.XGBoost

https://xgboost.readthedocs.io/en/latest/python/python_api.html

采用多棵CART树进行累加得到结果。每棵树可以随机选择特征与样本

16.AOC曲线和AUC值

https://www.cnblogs.com/dlml/p/4403482.html

17.使用sklearn进行增量学习

https://blog.youkuaiyun.com/whiterbear/article/details/53120004

https://www.zhihu.com/question/38034287?sort=created

12.plotly使用基本介绍

https://blog.youkuaiyun.com/tansuo17/article/details/79238505

13.Xgboost参数解释及调参

evals (list of pairs (DMatrix, string))：[(train,'train'),(test,'valid')]后面的string为返回结果时使用result['valid']获取结果

https://blog.youkuaiyun.com/iyuanshuo/article/details/80142730

14.深度学习教程

https://www.zybuluo.com/hanbingtao/note/433855

15.RNN详解

https://blog.youkuaiyun.com/wangyangzhizhou/article/details/76278375

https://blog.youkuaiyun.com/wangyangzhizhou/article/details/77679261

16.LSTM介绍

https://www.cnblogs.com/wushaogui/p/9176617.html

https://blog.youkuaiyun.com/qq_43631663/article/details/88563663

17 tensorflow rnn接口详解

https://zhuanlan.zhihu.com/p/28196873

18.rnn中time_step介绍

https://www.zhihu.com/question/271774530/answer/364711129

https://blog.youkuaiyun.com/weixin_37659245/article/details/89945487

19.为什么L1正则化比L2正则化更容易获得稀疏解

https://blog.youkuaiyun.com/Ahead_J/article/details/85100177

20.图神经网络

https://www.cnblogs.com/SivilTaram/p/graph_neural_network_1.html

21.使用Tensorflow训练LSTM+Attention中文标题党分类

https://blog.youkuaiyun.com/huanghaocs/article/details/85255227

22.决策树，随机森林可视化

https://blog.youkuaiyun.com/linhai1028/article/details/79827331

from sklearn.tree import export_graphviz
# Export as dot file
export_graphviz(estimator, out_file='tree.dot', 
                feature_names = iris.feature_names,
                class_names = iris.target_names,
                rounded = True, proportion = False, 
                precision = 2, filled = True)
 '''
# 使用系统命令转换为png（需要Graphviz）
from subprocess import call
call(['dot', '-Tpng', 'tree.dot', '-o', 'tree.png', '-Gdpi=600'])