Data Science & Machine Learning

Announcement

For appeals, questions and feedback about Oracle Forums, please email oracle-forums-moderators_us@oracle.com. Technical questions should be asked in the appropriate category. Thank you!

Model & Algorithm for fraudulent Credit Card transaction

735554Nov 18 2009 — edited Jan 25 2010

Hello Everyone,

I am new to datamining. Currently I am working on an acedemic project related to credit card fraudlent transactions. My objective is to classify a new transaction as normal or fraudlent. AFAIK the classification/scoring will be different for each card right? Like if some customer's past average transaction amount is let's say $100 and if a new transaction occurs from the same card whose transaction amount is let's say $1000 so that could be suspected as fraudlent transaction. So these conditions will vary to card to card or customer to customer. Please correct me if what i think is wrong?

Can anyone help me which model and algorithm is best for this scenerio? Any help would be greatly appreciated.

Locked Post

New comments cannot be posted to this locked post.

Locked on Feb 22 2010

Added on Nov 18 2009

#machine-learning

2 comments

2,414 views