Online-Payment-Fraud-Detection

This project is about to predict whether the payment done is fraud or not. Depending upon the features in the dataset, I have preidcted the output.

The libraries used in this model are as follows:

• Numpy
• Pandas
• Matplotlib.pyplot
• Sklearn

The dataset contains following features depending upon which I have predicted the output.

• Type of Payment
• Amount
• OldBalance
• NewBalance
• oldbalanceDest

1. I have imported the required libraries and Read the dataset.

2. Feature Engineering :

In this I have checked whether the dataset contains any null values or not , In this case there is no null value. If there is null values then I will replace with mean and median for contious and categorical features respectively.

3. Data Visualization:

Distributuion of transaction type is showed.

4. Now I have obtained the correlation between the features of the data with the isFraud column.

5. Divide the data into training and testing and trained the model using RandomForestClassifier algorithm

6. Now I have classified/predicted whether a transaction is a fraud or not by feeding about a transaction into the model

This is how I have built the Machine Learning Model which predicts the online payment is Fraud or Not.

Project by - Shivsharan Malage

Github

Linkedin