Data Shift : Detect Prior Probability Shift

[anilkumarpanda]

Problem Description Prior Probability Shift (PPS) refers the case when the input feature distributions remain the same but the distribution of the target variable changes.

Desired Outcome Since, probatus deals with binary classification only, the solution should :

• Plot the distribution of the target variables in the train and test set .
• Use a statistical test to identify PPS.

Solution Outline

• Plot the distribution of the target variables in the train and test set .
• Use Z test .

Refers to issue #72 comments