Coursera - linear regression - notes

cost function的表达式

 

batch gradient descent

1.**simultaneously update all theta**

2.每次都使用 所有都数据

 

Update equation的表达式,

就是求导之后的

 

这里,  X0 = 1

 

Intercept term (theta 0)

给X增加一列,全部为1, 把theta0也看成一个feature

 

画图

画红xxx

plot(x,y,'rx','MarkerSize',10);
ylabel('somethingYlable');
xlabel('xLable');

 

  画线

% 如果不用hold on 画布上之前的图就会消失
hold on; % keep previous plot visible
plot(X(:,2), X*theta, '-')
legend('Training data', 'Linear regression')
hold off % don't overlay any more plots on this figure

 

改良之后的 动态展示绘图 和 cost展示

    if mod(iter,200) == 0
        hold on; % keep previous plot visible
        plot(X(:,2), X*theta, '-');
        hold off % don't overlay any more plots on this figure
        current_J = computeCost(X, y, theta)
    end

  

plotPractice.m

matlab contour api

contour(Z,v) draws a contour plot of matrix Z with contour lines at the data values specified in the monotonically increasing vector v

contour(theta0_vals, theta1_vals, J_vals,[5 10 20 30 50 70 100 300])%logspace(-2, 3, 10)

 

figure; 
surf(theta0_vals, theta1_vals, J_vals)
xlabel('\theta_0'); ylabel('\theta_1');

 figure 是创建一个新图

theta0_vals 是 x

theta1_vals是 y

J_vals是 长为theta0 宽为 theta1的数组矩阵

 

logspace(-2, 3, 10) 

貌似是10的-2次方到3次方,10个间隔,

如果是(-5,5,11) 那就是严格按照10的x次方间隔

 

matlab矩阵

可以用逗号分隔,也可以用空格分隔

分号分隔column

 

standard deviation 标准差  均方差

标准差是方差的算术平方根

 

Feature Normalization

feature scaling

两步

  1.减去平均值

  2.除以标准差(均方差)

 

it’s helpful to plot J for several learning rates on the same figure.

with a ‘hold on’ command between plots 

可以在一张图上画多个cost曲线,来对比 learning rate

plot(1:numel(J_history), J_history, '-b', 'LineWidth', 2);

  x为训练次数,从1到numel, y为cost

numel: 计算元素个数

 

 

Normal Equations

no "loop until convergence"

exact solution of linear regression

 

 

% 如果不用hold on 画布上之前的图就会消失
hold on; % keep previous plot visible
plot(X(:,2), X*theta, '-')
legend('Training data', 'Linear regression')
hold off % don't overlay any more plots on this figure

  

posted @ 2017-07-18 18:20  烧鸭饭真好吃  阅读(229)  评论(0编辑  收藏  举报