# Payments Proactive Fraud Detector
## Summary
Detects fraudulent payment transactions for a retail customer based on his/her payment transactions history. It makes use of unsupervised learning algorithm oneClassSVM to be precise for detecting novelty of a payment transaction. It makes use different payment transaction attributes like amount, creditor, place of transaction, percentage of balance amount, time of transaction etc to detect novelty of current transaction.

## Tools Used
* Machine Learning Algorithm - oneClassSVM for n-dimensional data, pca, standardscalar using python libraries in scikit![Uploading file..._88rfaa3k7]()

## Installation Guide for Prototype
1. Procure Centos 7/8 VM with Python3.8 or 3.7 atleast.
2. Install virtualenv (assuming you used Centos8 in step 1)
``dnf install virtualenv
3.