Coursera吴恩达机器学习课程 总结笔记及作业代码——第1,2周

  • Linearregression
    • 1 Model representation
    • 2 Cost function
    • 3 Gradient descent
    • 4 Gradient descent for linear regression
    • 1 Mul2ple features
    • 2 Feature Scaling
    • 3 Learningrate
    • 4 Features and polynomial regression
    • 5 Normal equa2on
    • 编程作业

 

Linear’regression

发现这个教程是最入门的一个教程了,老师讲的很好,也很通俗,每堂课后面还有编程作业,全程用matlab编程,只需要填写核心代码,很适合自学。

1.1 Model representation

起始给出了预测房价的例子。 
这里写图片描述
这个问题属于监督问题,每个样本都给出了准确的答案。 
同时属于回归问题,对给定值预测实际输出。

定义(x(i),y(i))(x(i),y(i))为第i个样本,x表示输入值,y表示输出值,上标表示样本。

以下是机器学习运行模型 
这里写图片描述

对于假设h我们可以用一条直线描述,用线性函数预测房价值。 
hθ(x)=θ0+θ1∗xhθ(x)=θ0+θ1∗x

1.2 Cost function

我们取怎样的θθ值可以使预测值更加准确呢? 
想想看,我们应使得每一个预测值和真实值差别不大,可以定义代价函数如下 
J(θ0,θ1)=12m∑mi=1(hθ(x(i))−y(i))2J(θ0,θ1)=12m∑i=1m(hθ(x(i))−y(i))2 
通过使J值取最小来满足需求

下面通过图形方式感受一下代价函数 
这里写图片描述

1.3 Gradient descent

怎样使我们的代价函数取得最小值呢 
下面我们采取梯度下降法。 
这里写图片描述
好比我们下山,每次在一点环顾四周,往最陡峭的路向下走,用图形的方式更形象的表示 
这里写图片描述

Gradient descent algorithm 
repeat until convergence{ 
  θj=θj−α∂∂θjJ(θ0,θ1)θj=θj−α∂∂θjJ(θ0,θ1)  (for j=0 and j=1)(for j=0 and j=1) 
}

注意更新theta值应同时更新,matlab中向量更新即为同时更新,所以应使上式向量化(之后会讲解向量化含义),也可采取下面方式 
这里写图片描述

1.4 Gradient descent for linear regression

repeat until convergence{ 
  θj=θj−α∂∂θjJ(θ0,θ1)θj=θj−α∂∂θjJ(θ0,θ1)  (for j=0 and j=1)(for j=0 and j=1) 
}

 

∂∂θjJ(θ0,θ1)==∂∂θj12m∑i=1m(hθ(x(i)−y(i)))2∂∂θj12m∑i=1m(hθ(θ0+θ1x)−y(i))2∂∂θjJ(θ0,θ1)=∂∂θj12m∑i=1m(hθ(x(i)−y(i)))2=∂∂θj12m∑i=1m(hθ(θ0+θ1x)−y(i))2

 

j=0:∂∂θjJ(θ0,θ1)=1m∑mi=1(hθ(x(i)−y(i)))j=0:∂∂θjJ(θ0,θ1)=1m∑i=1m(hθ(x(i)−y(i))) 
j=1:∂∂θjJ(θ0,θ1)=1m∑mi=1(hθ(x(i)−y(i)))∗x(i)j=1:∂∂θjJ(θ0,θ1)=1m∑i=1m(hθ(x(i)−y(i)))∗x(i)

2.1 Mul2ple features

如果输入值不止一个,我们的假设函数应修改为 
hθ(x)=θ0+θ1x1+θ2x2+⋯+θnxnhθ(x)=θ0+θ1x1+θ2x2+⋯+θnxn

为了结构统一,我们设 x0=1x0=1 
hθ(x)=θ0+θ1x1+θ2x2+⋯+θnxn=θTxhθ(x)=θ0+θ1x1+θ2x2+⋯+θnxn=θTx 
这里写图片描述 
如此一来,便将变量向量化了

New algorithm 
repeat until convergence{ 
  θj=θj−α∂∂θjJ(θ)=θj−α1m∑mi=1(hθ(x(i)−y(i)))∗x(i)jθj=θj−α∂∂θjJ(θ)=θj−α1m∑i=1m(hθ(x(i)−y(i)))∗xj(i)  (for j=0,1,2⋯n)(for j=0,1,2⋯n) 
}

2.2 Feature Scaling

面对输入数据各个特征值范围差距过大的问题,我们可以对输入数据进行标准化。 
x(j)i=x(j)i−avg(xi)Sixi(j)=xi(j)−avg(xi)Si 
其中SiSi可以为标准差,也可以为max(xi)−min(xi)max(xi)−min(xi)

2.3 Learning’rate

  1. 如果αα太小,则梯度下降法会收敛缓慢
  2. 如果αα太大,则梯度下降法每次迭代可能不下降,最终导致不收敛。

2.4 Features and polynomial regression

除了线性回归外,我们也能采用多项式回归 
举例如下假设函数 
hθ(x)=θ0+θ1x+θ2x2+θ3x3hθ(x)=θ0+θ1x+θ2x2+θ3x3 
我们可以定义为 
hθ(x)=θ0+θ1x1+θ2x2+θ3x3=θ0+θ1x1+θ2x21+θ3x31hθ(x)=θ0+θ1x1+θ2x2+θ3x3=θ0+θ1x1+θ2x12+θ3x13 
对于多项式回归,标准化更加重要。

2.5 Normal equa2on

除了梯度下降法,另一种求最小值的方式则是让代价函数导数为0,求θθ值 
J(θ)=12m∑mi=1(hθ(x(i))−y(i))2J(θ)=12m∑i=1m(hθ(x(i))−y(i))2 
∂∂θjJ(θ)=0∂∂θjJ(θ)=0 for every j 
求得: θ=(XTX)−1XTyθ=(XTX)−1XTy

下面这个图比较了两个算法之间的区别 
这里写图片描述

 

特殊情况:

        由于用标准方程法时,涉及到要计算矩阵XTX的逆矩阵。但是XTX的结果有可能不可逆。    当使用python的numpy计算时,其会返回广义的逆结果。    主要原因:    出现这种情况的主要原因,主要有特征值数量多于训练集个数、特征值之间线性相关(如表示面积采用平方米和平方公里同时出现在特征值中)。    因此,首先需要考虑特征值是否冗余,并且清除不常用、区分度不大的特征值。对于 (XTX)(XTX)不可逆的情况下,我们可以采取减少特征量和使用正规化方式来改善。
比较标准方程法和梯度下降法:

 这两个方法都是旨在获取使代价函数值最小的参数θ,两个方法各有优缺点:  

       1)梯度下降算法   

              优点:当训练集很大的时候(百万级),速度很快。    

               缺点:需要调试出合适的学习速率α、需要多次迭代、特征值数量级不一致时需要特征缩放。

       2)标准方程法:

                优点:不需要α、不需要迭代、不需要特征缩放,直接解出结果。    

                缺点:运算量大,当训练集很大时速度非常慢。
综合:因此,当训练集百万级时,考虑使用梯度下降算法;训练集在万级别时,考虑使用标准方程法。在万到百万级区间时,看情况使用,主要还是使用梯度下降算法。

编程作业

ex1.m

%% Machine Learning Online Class - Exercise 1: Linear Regression

%  Instructions
%  ------------
%
%  This file contains code that helps you get started on the
%  linear exercise. You will need to complete the following functions
%  in this exericse:
%
%     warmUpExercise.m
%     plotData.m
%     gradientDescent.m
%     computeCost.m
%     gradientDescentMulti.m
%     computeCostMulti.m
%     featureNormalize.m
%     normalEqn.m
%
%  For this exercise, you will not need to change any code in this file,
%  or any other files other than those mentioned above.
%
% x refers to the population size in 10,000s
% y refers to the profit in $10,000s
%

%% Initialization
clear ; close all; clc

%% ==================== Part 1: Basic Function ====================
% Complete warmUpExercise.m
fprintf('Running warmUpExercise ... \n');
fprintf('5x5 Identity Matrix: \n');
warmUpExercise()

fprintf('Program paused. Press enter to continue.\n');
pause;


%% ======================= Part 2: Plotting =======================
fprintf('Plotting Data ...\n')
data = load('ex1data1.txt');
X = data(:, 1); y = data(:, 2);
m = length(y); % number of training examples

% Plot Data
% Note: You have to complete the code in plotData.m
plotData(X, y);

fprintf('Program paused. Press enter to continue.\n');
pause;

%% =================== Part 3: Cost and Gradient descent ===================

X = [ones(m, 1), data(:,1)]; % Add a column of ones to x
theta = zeros(2, 1); % initialize fitting parameters

% Some gradient descent settings
iterations = 1500;
alpha = 0.01;

fprintf('\nTesting the cost function ...\n')
% compute and display initial cost
J = computeCost(X, y, theta);
fprintf('With theta = [0 ; 0]\nCost computed = %f\n', J);
fprintf('Expected cost value (approx) 32.07\n');

% further testing of the cost function
J = computeCost(X, y, [-1 ; 2]);
fprintf('\nWith theta = [-1 ; 2]\nCost computed = %f\n', J);
fprintf('Expected cost value (approx) 54.24\n');

fprintf('Program paused. Press enter to continue.\n');
pause;

fprintf('\nRunning Gradient Descent ...\n')
% run gradient descent
theta = gradientDescent(X, y, theta, alpha, iterations);

% print theta to screen
fprintf('Theta found by gradient descent:\n');
fprintf('%f\n', theta);
fprintf('Expected theta values (approx)\n');
fprintf(' -3.6303\n  1.1664\n\n');

% Plot the linear fit
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

% Predict values for population sizes of 35,000 and 70,000
predict1 = [1, 3.5] *theta;
fprintf('For population = 35,000, we predict a profit of %f\n',...
    predict1*10000);
predict2 = [1, 7] * theta;
fprintf('For population = 70,000, we predict a profit of %f\n',...
    predict2*10000);

fprintf('Program paused. Press enter to continue.\n');
pause;

%% ============= Part 4: Visualizing J(theta_0, theta_1) =============
fprintf('Visualizing J(theta_0, theta_1) ...\n')

% Grid over which we will calculate J
theta0_vals = linspace(-10, 10, 100);
theta1_vals = linspace(-1, 4, 100);

% initialize J_vals to a matrix of 0's
J_vals = zeros(length(theta0_vals), length(theta1_vals));

% Fill out J_vals
for i = 1:length(theta0_vals)
    for j = 1:length(theta1_vals)
      t = [theta0_vals(i); theta1_vals(j)];
      J_vals(i,j) = computeCost(X, y, t);
    end
end


% Because of the way meshgrids work in the surf command, we need to
% transpose J_vals before calling surf, or else the axes will be flipped
J_vals = J_vals';
% Surface plot
figure;
surf(theta0_vals, theta1_vals, J_vals)
xlabel('\theta_0'); ylabel('\theta_1');

% Contour plot
figure;
% Plot J_vals as 15 contours spaced logarithmically between 0.01 and 100
contour(theta0_vals, theta1_vals, J_vals, logspace(-2, 3, 20))
xlabel('\theta_0'); ylabel('\theta_1');
hold on;
plot(theta(1), theta(2), 'rx', 'MarkerSize', 10, 'LineWidth', 2);

ComputeCost.m

gradientDescent.m

function [theta, J_history] = gradientDescent(X, y, theta, alpha, num_iters)
%GRADIENTDESCENT Performs gradient descent to learn theta
%   theta = GRADIENTDESCENT(X, y, theta, alpha, num_iters) updates theta by 
%   taking num_iters gradient steps with learning rate alpha

% Initialize some useful values
m = length(y); % number of training examples
J_history = zeros(num_iters, 1);

for iter = 1:num_iters

    % ====================== YOUR CODE HERE ======================
    % Instructions: Perform a single gradient step on the parameter vector
    %               theta. 
    %
    % Hint: While debugging, it can be useful to print out the values
    %       of the cost function (computeCost) and gradient here.
    %

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

    % ============================================================

    % Save the cost J in every iteration    
    J_history(iter) = computeCost(X, y, theta);

end

end
  • 这里写图片描述

这里写图片描述

你可能感兴趣的:(机器学习,机器学习)