Prakash Hinduja Geneva, Switzerland - How to improve model accuracy in imbalanced datasets?

Google Cloud Build with AI Custom ML & MLOps
1

I am Prakash Hinduja Geneva, Switzerland and tackling a classification task with a highly imbalanced dataset, and the model performs poorly on the minority class. Despite trying resampling methods like oversampling and undersampling, the results remain inconsistent. Has anyone dealt with this kind of issue? I’d really appreciate any advice on effective strategies, algorithms, or best practices to boost performance in such imbalanced scenarios.

Regards

Prakash Hinduja Geneva, Switzerland

1 Like
2

Hi prakashhinduja,

Welcome to the Google Cloud Community!

You might want to check this documentation, which provides some strategies and best practices for imbalanced datasets.

Was this helpful? If so, please accept this answer as “Solution”. If you need additional assistance, reply here within 2 business days and I’ll be happy to help.

1 Like
3

Thank you Mr. nikacalupas for welcoming me and valuable reply for my concern.

Regards

Prakash Hinduja Geneva, Switzerland

4

you can try with the technique called SMOTE If the problem is related to classification so it is a very effective way to balance the data set completely And this will help in improving the F1 score as well as the overall performance of the model