I am currently working on a dataset which is imbalanced. It has 2850 negative label points and 483 positive label points and I do upsample on the dataset to balance it. But my model performance decreases after I feed upsampled data to my machine learning model.

For example : When I feed the imbalance data to xgboost .The performance metrics are : Precision = 0.991 , Recall = 0.504 , fscore = 0.669 , Accuracy =0.933

and When I feed balanced data to xgboost .The performance metrics changes to : Precision = 0.401 , Recall = 0.656 , fscore = 0.497 , Accuracy =0.822

Other models like Logistic regression, Naive Bayes, KNN performance degrades drastically after feeding them the upsampled data.

My question is, can you let me understand the behavior of the models? Why does the performance decreases drastically when I feed upsampled data to the machine learning models?

Your Answer

 
discard

By clicking "Post Your Answer", you acknowledge that you have read our updated terms of service, privacy policy and cookie policy, and that your continued use of the website is subject to these policies.

Browse other questions tagged or ask your own question.