Principle analysis and code implementation of logistic regression classification algorithm

Summary

1. The computational cost of logistic regression is not high, it is a very common classification algorithm. The centralized logistic regression classifier based on random gradient rise can support online learning.

2. However, the disadvantage of the logistic regression algorithm is obvious-it can only solve the classification problem of two classes.

3. In addition, the logical regression is easy to fit, resulting in a low accuracy of classification.

Principle analysis and code implementation of logistic regression classification algorithm