Feature selection: maximum information coefficient (MIC) [used to measure the degree of correlation between two variables X and y, linear or nonlinear strength, commonly used for feature selection of

One 、 Concept

1.1 MIC
MIC, namely （Maximal Information Coefficient） Maximum information coefficient , Belong to Maximal Information-based Nonparametric Exploration (MINE) The largest nonparametric exploration based on information , Used to measure two variables X and Y The degree of correlation between , Linear or nonlinear strength , Feature selection commonly used in machine learning .

    MIC Compare with Mutual Information（MI） Higher accuracy in terms of mutual information , It is an excellent calculation method of data relevance .

1.2 Mutual information
Mutual information （Mutual Information） It's a useful measure of information in information theory , It can be seen as the amount of information about another random variable contained in one random variable , Or the uncertainty of a random variable reduced by knowing another random variable . This is already a cliche in machine learning , If I can't remember , Please refer to Baidu Encyclopedia - Mutual information .

---

---

---

Reference material ：
MIC： Maximum information coefficient