An Intelligent Online Payment Fraud Detection System Using Machine Learning and Explainable AI

Abstract

With the rapid growth of digital transactions and online payment systems, financial fraud has
become a major concern for individuals, businesses, and financial institutions. Fraudulent
activities such as unauthorized transactions, identity theft, and payment manipulation result in
significant financial losses and undermine trust in digital platforms. Traditional fraud detection
methods, which rely on rule-based systems, are often inefficient in identifying complex and
evolving fraud patterns. This project presents an intelligent Online Payment Fraud Detection
System that leverages machine learning algorithms combined with explainable artificial
intelligence (XAI) to improve detection accuracy and transparency.
The proposed system utilizes three widely used classification algorithms: Logistic Regression,
Random Forest, and Support Vector Machine (SVM). These models are trained on transactional
datasets to identify patterns indicative of fraudulent activities. The system incorporates a
preprocessing pipeline that converts categorical variables into numerical form using one-hot
encoding and handles missing values using imputation techniques. The dataset is then split into
training and testing sets to ensure reliable performance evaluation.
A key feature of the system is the use of a pipeline-based architecture, which integrates
preprocessing and model training into a single workflow. This ensures consistency and reduces
the risk of data leakage. Model performance is evaluated using accuracy metrics, and a
comparison module allows users to analyze the effectiveness of different algorithms.
Additionally, the system includes ROC curve visualization to assess classification performance in
terms of true positive and false positive rates