Power Platform Community Forum Thread Details

My custom model predicts whether a transaction is fraudulent or not.

It has been trained using this dataset.

The accuracy score is 99% on the training dataset (probably overfitting...).

Now, when looking at the generated prediction columns, after running the model, I get a lot of empty values:

Here you may see the label and the prediction column:

While some other rows have values:

The table contains 1.994.188 rows.

Categories:

Binary classification model does not predict on every rows