[人工智能]神经网络优化方向及模型不收敛的原因

最近在使用MLP做一个数据预测， 在调整模型的同时顺便做了个思维导图来记录下相关知识

由于博主平常主要用的是基于TF 1.x的keras，所以有些内容多少有些偏向

参考链接：

1. https://keras.io/zh/
2. https://zhuanlan.zhihu.com/p/36369878
3. https://blog.csdn.net/promisejia/article/details/88801307
4. https://zhuanlan.zhihu.com/p/390990848
5. https://zhuanlan.zhihu.com/p/172254089
6. https://zhuanlan.zhihu.com/p/390261440
7. https://www.cnblogs.com/shine-lee/p/11908610.html#scroller-6
8. https://www.nowcoder.com/discuss/421969?type=6
9. https://www.jianshu.com/p/f5049e8a27d0
10. https://www.jianshu.com/p/aebcaf8af76e
11. https://zhuanlan.zhihu.com/p/109521635
12. https://www.jiqizhixin.com/graph/technologies/26dfa740-8671-4c05-8ed3-c678b9335b4a
13. https://blog.csdn.net/qq_42109740/article/details/105401197
14. https://blog.csdn.net/tototuzuoquan/article/details/112724028
15. https://blog.csdn.net/weixin_44023658/article/details/105844861
16. https://blog.csdn.net/fengbingchun/article/details/124909910?spm=1001.2101.3001.6650.1&utm_medium=distribute.pc_relevant.none-task-blog-2%7Edefault%7ECTRLIST%7ERate-1-124909910-blog-112094166.pc_relevant_aa&depth_1-utm_source=distribute.pc_relevant.none-task-blog-2%7Edefault%7ECTRLIST%7ERate-1-124909910-blog-112094166.pc_relevant_aa&utm_relevant_index=2

需要导图源文件和相关导图软件的可以留言