This is a course project for EE 401 Pattern Recognition and Machine Learning, done in collaboration with three of my classmates.
About
Credit card fraud can be defined as, ‘Unauthorized account activity involving a payment card, by a person for which the account is not intended’. These frauds cost consumers and banks millions of dollars worldwide, as a response to which several modern fraud-detection techniques are in place today.
Data ResourcesWe are working on the dataset provided in Kaggle, which is available under Open Database license.
Proposed Algorithm
We decompose the proposed algorithm into three parts:
• Bagging of feature vectors
• Assigning weights to the classifier associated with every bag
• Obtaining the model for classification
• Testing the dataset with obtained classifier
Parallel and incremental credit card fraud detection model to handle concept drift and data imbalance
Somasundaram, A. & Reddy, S.
Neural Computing and Applications, Springer, (2019) 31(Suppl 1): 3
Fraud Detection using Machine Learning
Aditya Oza, Stanford University