1、对特征选择和模型融合的一些建议
http://www.cnblogs.com/zhizhan/p/5826089.html
http://blog.csdn.net/wtq1993/article/details/51418958
2、特征工程
http://www.cnblogs.com/zhizhan/p/5826089.html
https://www.zhihu.com/question/28641663?sort=created
3、KDD2017
https://yq.aliyun.com/articles/205495
4、tips
（1）xgb中自定义metric
http://www.cnblogs.com/silence-gtx/p/5812012.html
（2）模型集成
http://blog.csdn.net/m0_37728157/article/details/74572239
（3）特征选择
http://www.cnblogs.com/hhh5460/p/5186226.html
http://scikit-learn.org/stable/modules/feature_selection.html#feature-selection
http://www.cnblogs.com/harvey888/p/5852757.html
（4）kaggle建议
https://www.leiphone.com/news/201705/xBIkv30NVrMqdMdX.html
（5）FM、FFM
http://blog.csdn.net/bitcarmanlee/article/details/52143909
http://blog.csdn.net/mmc2015/article/details/51760681
https://tech.meituan.com/deep-understanding-of-ffm-principles-and-practices.html
（6）ctr预估
http://blog.csdn.net/lilyth_lilyth/article/details/48032119
（7）分类算法
http://blog.csdn.net/jackywu1010/article/details/7055561
（8）gbdt
http://blog.csdn.net/zhangf666/article/details/70174464
（9）bagging boosting
http://blog.csdn.net/jlei_apple/article/details/8168856
（10）DNN ctr预估
http://blog.csdn.net/xiewenbo/article/details/65446549?locationNum=15&fps=1
5、tensorflow预测实例
http://www.jianshu.com/p/9a00ff620e6e
6、梯度下降
http://www.cnblogs.com/maybe2030/p/5089753.html
http://blog.csdn.net/itplus/article/details/21896453
http://www.cnblogs.com/downtjs/p/3222643.html
7、alibaba
https://www.leiphone.com/news/201707/t0AT4sIgyWS2QWVU.html
8、tensorflow
http://www.cnblogs.com/wuzhitj/p/6298011.html
9、word2vec
http://techblog.youdao.com/?p=915#LinkTarget_699
http://nooverfit.com/wp/pycon-2016-tensorflow-%E7%A0%94%E8%AE%A8%E4%BC%9A%E6%80%BB%E7%BB%93-tensorflow-%E6%89%8B%E6%8A%8A%E6%89%8B%E5%85%A5%E9%97%A8-%E7%AC%AC%E4%BA%8C%E8%AE%B2-word2vec/
10、maxnet
http://blog.csdn.net/u012556077/article/details/50409842
11、机器学习 人工智能
http://blog.csdn.net/pongba/article/details/2915005
12、
https://me.csdn.net/v_JULY_v
https://blog.csdn.net/v_JULY_v/article/details/89894058
