AD CLICK FRAUD DETECTION

Abstract

Ad click fraud has become one of the most critical challenges in the digital advertising ecosystem, causing huge financial losses to advertisers and reducing trust in online marketing platforms. Fraudulent clicks are generated either manually or through automated bots with the intention of inflating advertisement revenue, exhausting competitor budgets, or manipulating analytics. Traditional rule-based systems are no longer sufficient to handle the scale and complexity of modern fraud patterns. Therefore, this research proposes a datadriven approach using machine learning and deep learning algorithms for detecting and analyzing ad click fraud.The study explores the performance of multiple algorithms such as Convolutional Neural Networks (CNN), Artificial Neural Networks (ANN), Recurrent Neural Networks (RNN), Random Forest, and a proposed XGBoost-based system. Realworld ad click datasets are preprocessed through cleaning, normalization, feature engineering, and feature selection. Models are trained and evaluated using standard metrics such as accuracy, precision, recall, F1-score, and AUC. The results show that ensemblebased methods, especially XGBoost, outperform traditional and deep learning models in terms of accuracy, interpretability, and computational efficiency. The research concludes that combining robust preprocessing with advanced machine learning models can significantly improve fraud detection and reduce financial losses in digital advertising.