Skip to content

Latest commit

 

History

History
62 lines (42 loc) · 1.89 KB

File metadata and controls

62 lines (42 loc) · 1.89 KB

Fraud Credit Card Transaction Classification

Download Dataset from kaggle-fraud-transaction-classification

In the project, we will be working on machine learning project on credit card fraud


Table Of Contents


Project Structure Overview

├── Fraud Credit Card Transaction Classification
|  ├── data          - this folder contains training & testing data.
|  |    └──  transactions.csv (Ommited)
|  │
└────── Fraud Transaction Classification.ipynb

Project Goal

The following steps are followed:

  • Describe the structure of the date (statistical summary: number of records, counting max, min, null,...)
  • Plotting histogram of transactionAmount column.
  • We have duplicated transactions in the dataset - two types: reversed transaction and multi-swipe. You need to answer these questions:
    • Identify those types of duplicates
    • Estimate of the total number of transactions and total dollar amount

How do you estimate for the reversed transactions and multi-swipe transactions?

  • The first transaction to be normal and exclude it from the number of transaction and dollar amount count

  • Modeling - Provide modeling algorithms and say why you use those methods? Why used those features?

    • There is a type of transactions in the dataset has a field called isFraud. Build a predictive model to determine whether a given transaction will be fraudulent or not.
    • Use an estimate of performance using an appropriate sample

Version

1.0.0


Author

  • Rahul Gaikwad - Initial work and development

I welcome your questions. Write to rahul.gaikwad2010@gmail.com