Abstract

The rapid growth of digital healthcare systems has enhanced the risk of fraudulent activities in health insurance claims, leading to financial losses and inefficient resource allocation. Traditional fraud detection methods often fail to identify complex patterns in large healthcare datasets. This study proposes a fraud detection framework that integrates machine learning and deep learning techniques to analyse structured claim data and sequential patient records. Models such as Logistic Regression, Random Forest, Convolutional Neural Networks (CNN), and Long Short-Term Memory (LSTM) are used to detect fraudulent claims. The system incorporates data preprocessing, feature engineering, and evaluation metrics to improve detection accuracy while ensuring data privacy and regulatory compliance.