Question

我对机器学习非常陌生。想用Coursera的课程训练自己。我正在使用Logistic回归进行＆＃34; winequality-white＆＃34;数据。数据样本如下

7   0.27    0.36    20.7    0.045   45  170 1.001   3       0.45    8.8 6
6.3 0.3     0.34    1.6     0.049   14  132 0.994   3.3     0.49    9.5 6
8.1 0.28    0.4     6.9     0.05    30  97  0.9951  3.26    0.44    10  6
7.2 0.23    0.32    8.5     0.058   47  186 0.9956  3.19    0.4     9.9 6
7.2 0.23    0.32    8.5     0.058   47  186 0.9956  3.19    0.4     9.9 6

值为6的最后一列实际上包含3到9的值表示葡萄酒的质量。列的其余部分包含决定葡萄酒质量的各种参数。

根据课程，我确实构建了返回J和Gradient的Cost函数。如下

function [J, grad] = lrCostFunction(theta, X, y, lambda)

  m = length(y); 
  J = 0;
  grad = zeros(size(theta));
  h = sigmoid(X * theta); % hypothisis for logistic regression
  J = (1/m) * sum( (-y .* log(h)) - ((1-y) .* log(1-h))) + (lambda/(2*m)) * sum(theta(2:length(theta)).^2);
  grad(1) = (1 / m) * ( X'(1,:)) * (h-y);
  grad(2:size(theta,1)) = (1/ m) * (X'(2:size(X',1),:)*(h-y) + ...
                        lambda*theta(2:size(theta,1),:));
  grad = grad(:);
end

我从另一个函数调用此成本函数来查找Theta值

function [all_theta] = oneVsAll(X, y, num_labels, lambda)

% Some useful variables
m = size(X, 1);
n = size(X, 2);

% You need to return the following variables correctly 
all_theta = zeros(num_labels, n + 1);

% Add ones to the X data matrix
X = [ones(m, 1) X];

     for c = 1:num_labels  

%     % Set Initial theta
     initial_theta = zeros(n + 1, 1);
%     
%     % Set options for fminunc
     options = optimset('GradObj', 'on', 'MaxIter', 1000);
%     
      adc = c+2  ;
%     % Run fmincg to obtain the optimal theta
%     % This function will return theta and the cost 
     [theta] = ...
         fmincg (@(t)(lrCostFunction(t, X, (y == adc), lambda)), ...
                 initial_theta, options);

                 all_theta(c,:) = theta';

      end
end

在all_theta的帮助下，我正在预测值（我确实使用了训练和验证集）预测完全没有问题。有人可以帮助我在哪里犯错误。

这是我的第一个ML模型和有趣的部分，它给出准确率2％:) ..请帮助我朝正确的方向前进。提前致谢

注意： - 我正在使用Octave。

我也尝试使用Lambda值为0.1,0.01,0.001等。但没有运气:(

添加预测代码---

function p = predictOneVsAll(all_theta, X)

m = size(X, 1);
num_labels = size(all_theta, 1);

% You need to return the following variables correctly 
p = zeros(size(X, 1), 1);

% Add ones to the X data matrix
X = [ones(m, 1) X];

z = X * all_theta';
A = sigmoid(z);
[t,p]=max(A,[],2);
end

Logistic回归 - 在拟合下

0 个答案: