【机器学习】1.从梯度下降法开始

 

符号说明：

m:#training examples

x:input variables/features

y:output variable/target

(x,y):training example

i^th training example:(x⁽ⁱ⁾,y⁽ⁱ⁾)

 

梯度下降法

https://blog.csdn.net/qq_41800366/article/details/86583789

https://www.zhihu.com/question/305638940

批梯度下降法等衍生方法

https://blog.csdn.net/cs24k1993/article/details/79120579

正规方程组，梯度下降法的总结

注：用到的线性代数知识

trAB=trBA

类似地trABC=...

trA=trA^T

a∈R，则a=tr a

▽_AtrABA^TC=CAB+C^TAB^T

https://blog.csdn.net/hahajinbu/article/details/49904665

过拟合与欠拟合

https://blog.csdn.net/xuaho0907/article/details/88649141

逻辑回归

https://www.cnblogs.com/samsons/p/4419190.html

退火算法

https://www.cnblogs.com/heaad/archive/2010/12/20/1911614.html

牛顿方法

https://blog.csdn.net/qq_36330643/article/details/78003952

指示器

https://www.cnblogs.com/jmzz/archive/2011/08/11/2135073.html

感知器算法