For data scientists checking a classification dataset before training. You get the number of classes, total samples, the largest to smallest imbalance ratio, the majority baseline accuracy and each class share.
Build industry projects on ByteLabs and add verified proof of your skills to your public profile.
You enter
cat: 900, dog: 90, bird: 10
The tool shows
3 classes, total 1,000, imbalance ratio 90 : 1, majority baseline 90%; shares 90%, 9%, 1%
It is the accuracy you get by always predicting the largest class. Here 900 of 1,000 samples gives 90 percent, so a model must beat that to add value.
There is no fixed cutoff, but ratios beyond about 10 to 1 usually need attention, and 100 to 1 or more calls for metrics like F1 or PR-AUC instead of accuracy.
Common options are class weights in the loss, resampling (oversampling the minority or undersampling the majority), synthetic samples such as SMOTE, and threshold tuning.
Yes. It is free, needs no sign-up and runs entirely in your browser, so what you type is not uploaded. You only sign in if you want to email a result to yourself or save it to your CareerByteCode profile.
