【Stanford Machine Learning Open Course】6. week2编程题解

这里是斯坦福大学机器学习网络课程的学习笔记。课程地址是：https://class.coursera.org/ml-2012-002/lecture/index

 

 

编程题地址：https://class.coursera.org/ml-2012-002/assignment/index

题目包括计算成本函数，下降梯度， 正规方程组，特征归一化

 

%1/7 warm up:

A=eye(5)

 

%2/7, 5/7 compute cost function (for one or multi variables) 

J=1/(2*m)*sum((X*theta-y).^2);

 

%3/7, 6/7Gradient Descent (for one or multi variables)

theta = theta-alpha/m*((X*theta-y)'*X)';

 

%4/7 Feature Normalization

mu=mean(X);

sigma=std(X);

X_norm = X - ones(size(X)*diag(mu)) ./ (ones(size(X)*diag(sigma))

 

%7/7 Normal Equation

theta=pinv(X'*X)*X'*y;