I am trying to get idea how variables of my data correspond to target variable (binary class).

In regression, Pearson r statistic is quite good to get sense of variable relationship. Also I can use it for classification, treating classes 0 and 1 as real values but it's a risky trick.

My question: is there any equivalent statistic saying which variable is good for classification task ? Thanks


The Pearson Chi-Squared test may give you what you need: https://en.wikipedia.org/wiki/Pearson%27s_chi-squared_test

This is a good resource for analyzing categorical data: https://en.wikipedia.org/wiki/List_of_analyses_of_categorical_data

Also, confusion matrices can be very helpful: https://en.wikipedia.org/wiki/Confusion_matrix

| improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.