Logistic regression and linear regression

First Linear regression Both the definition field and the value field of are from negative infinity to positive infinity , It is used to predict a value , For example, knowing height predicts weight .

Logistic regression is actually evolved from linear regression

Logistic regression is generally used to do binary classification , That is, the sum of probability and 0.5 compare , The probability is greater than 0.5 Belong to one class , Others fall into another category . So I hope the value range is （0,1）, So there is a logic function . Add the linear regression model and you have a logistic regression .

The decision boundary of logistic regression is linear

It can be inferred from the above formula , The decision boundary is a straight line .（ The probability of all elements on the line is 0.5）

in other words , First, we can get the equation of decision boundary , From the equation, we can know whether the decision boundary is linear or nonlinear .