收藏一些链接

怎么把方程形式化成矩阵：y=xtAx
https://blog.csdn.net/sunbobosun56801/article/details/103299602
最小主子式判断是否正定：
https://www.dr-eriksen.no/teaching/GRA6035/2010/lecture5.pdf
用代数余子式求逆矩阵：https://blog.csdn.net/daduzimama/article/details/120509856?spm=1001.2101.3001.6661.1&utm_medium=distribute.pc_relevant_t0.none-task-blog-2%7Edefault%7ECTRLIST%7ERate-1.pc_relevant_antiscanv2&depth_1-utm_source=distribute.pc_relevant_t0.none-task-blog-2%7Edefault%7ECTRLIST%7ERate-1.pc_relevant_antiscanv2&utm_relevant_index=1
代数余子式——转置得到伴随矩阵A*，适用于三行三列以上的矩阵。
PCA计算https://blog.csdn.net/qq_41598072/article/details/85009130

SVM流程：
https://blog.csdn.net/myarrow/article/details/51261971；https://www.zybuluo.com/tianxingjian/note/1755051
EDA探索性数据分析介绍：https://www.cnblogs.com/huzihu/p/11146493.html
机器学习建模流程：https://www.jianshu.com/p/59e759ba6ee4
用于特征选择的F检验和互信息法：https://www.icode9.com/content-4-466512.html
不同特征选择模型的区别：https://www.cnblogs.com/hhh5460/p/5186226.html；https://cloud.tencent.com/developer/article/1556213
激活函数定义及公式https://zhuanlan.zhihu.com/p/48776056
python合并表格各类函数的区别：https://www.jianshu.com/p/a06279b18ab8

深度学习概念及电子书
https://stanford.edu/~shervine/teaching/cs-230/cheatsheet-convolutional-neural-networks
中文https://zh-v1.d2l.ai/chapter_convolutional-neural-networks/index.html
英文https://d2l.ai/chapter_deep-learning-computation/index.html
各类神经网络区别与联系
计算速度取决与卷积层，存储空间取决于全连接层（参数个数）
https://www.its203.com/article/qq_26249811/104378703#LeNet
https://www.jiqizhixin.com/articles/2019-06-17-14
https://icode.best/i/36241747076183
RNN前向传播和反向传播
https://towardsdatascience.com/recurrent-neural-networks-rnns-3f06d7653a85