lksfr/credit_card_fraud

This repository includes the code used in my corresponding Medium post.

credit_card_fraud's Issues

Why to use two StandardScaler?

Hi,

I didnt understand why you used two standardscaler to scale Amount and Time features? I mean why you didnt use one standardscaler as common? Is there a special reason for this?

Edit: I didnt want to open another issue for ask another quick question. And it is that why you split train and test with 0.9/0.1 rate and you splitted train data again with the train_test_split method?

What is the final aim of the project?

Hi! A machine learning beginner here. I read your medium blog post and greatly appreciated the step by step procedure you described there. However I'm still confused as to the aim of the project. Since the model does not operate on any new data, it isn't a predictive analysis project. Since the model already has access to labelled data, what exactly is the model detecting?