curiousily / credit-card-fraud-detection-using-autoencoders-in-keras

iPython notebook and pre-trained model that shows how to build deep Autoencoder in Keras for Anomaly Detection in credit card transactions data

Credit Card Fraud Detection using Autoencoders in Keras

Full explanation can be found in this blog post. The source code is compatible with TensorFlow 1.1 and Keras 2.0.4

Hands-On Machine Learning from Scratch

Interested in deeper understanding of Machine Learning algorithms? Implement them in Python from scratch:

Read the book here

credit-card-fraud-detection-using-autoencoders-in-keras's Issues

the raw data and how to pre-process

Hi,
I got your data now, but i have a question, could you please share the raw-data or show the way how you process the data? Thank you very much.

Regards!

Why are you using relu on the last layer?

input_layer = Input(shape=(input_dim, ))

encoder = Dense(encoding_dim, activation="tanh", 
                activity_regularizer=regularizers.l1(10e-5))(input_layer)
encoder = Dense(int(encoding_dim / 2), activation="relu")(encoder)

decoder = Dense(int(encoding_dim / 2), activation='tanh')(encoder)
decoder = Dense(input_dim, activation='relu')(decoder)

autoencoder = Model(inputs=input_layer, outputs=decoder)

From the fraud_detection.ipynb, there's model using relu as last layer. However, the csv file contains negative values which relu cannot represent. I think the last layer of decoder should represent the input value. Wouldn't it be an issue?

Thanks.

how to define encoding_dim for 21 columns?

Why did we use the threshold of 2.9