How Logistic Regression (logistic regression) model to achieve binary (why do classification)

Here Insert Picture Description
1. First assume function is defined as h _ {\ theta} (x) = \ theta ^ {T} x at this time is assumed to be a function of the range (- \ infty, \ infty).
Here Insert Picture Description
Here Insert Picture Description

Solving the decision boundary

Here Insert Picture Description

Here Insert Picture Description

The introduction of cost function

Since the purpose of cost function is to find the true value of all samples - a minimum sum of squares of the predicted values, but due to the presence logistic regression sigmoid function, such that the normal function of the cost can not be lowered with a gradient method for solving the optimal solution, so according to the sigmoid function, to adjust the cost function, to obtain the following equation
Here Insert Picture Description
Here Insert Picture Description
Here Insert Picture Description
Here Insert Picture Description
Here Insert Picture Description

Published 49 original articles · won praise 13 · views 20000 +

Guess you like

Origin blog.csdn.net/weixin_44166997/article/details/90723529