I think you are facing an unbalanced dataset problem (fraud), and here are some Q&A associated with it.

Poor performance shown on Rare event modeling

https://stats.stackexchange.com/questions/134599/how-to-classify-a-unbalanced-dataset-by-convolutional-neural-networks-cnn

I think you are facing an unbalanced dataset problem (fraud), and here are some Q&A associated with it.

Poor performance shown on Rare event modeling

https://stats.stackexchange.com/questions/134599/how-to-classify-a-unbalanced-dataset-by-convolutional-neural-networks-cnn

I think you are facing an unbalanced dataset problem (fraud), and here are some Q&A associated with it.

Poor performance shown on Rare event modeling

http://stats.stackexchange.com/questions/134599/how-to-classify-a-unbalanced-dataset-by-convolutional-neural-networks-cnn

I think you are facing an unbalanced dataset problem (fraud), and here are some Q&A associated with it.

Poor performance shown on Rare event modeling

https://stats.stackexchange.com/questions/134599/how-to-classify-a-unbalanced-dataset-by-convolutional-neural-networks-cnn