MACHINE LEARNING AND DEEP NEURAL NETWORK APPROACH FOR ONLINE AD CLICK FRAUD DETECTION

Abstract

Online advertising has become one of the most important revenue sources for digital platforms, but it is increasingly threatened by ad click fraud, where malicious users or automated bots generate fake clicks on advertisements to manipulate advertising costs and revenues. These fraudulent activities lead to significant financial losses for advertisers and reduce the reliability of online advertising ecosystems. To address this challenge, this study proposes a Machine Learning and Deep Neural Network– based approach for detecting online ad click fraud. The proposed system utilizes historical clickstream data and user behavioral features such as IP address patterns, click frequency, session duration, device information, and geographic location to identify suspicious activities. Initially, data preprocessing and feature extraction techniques are applied to clean and transform raw click data into meaningful inputs for predictive models. Machine learning algorithms such as Random Forest, Decision Tree, and Logistic Regression are used to classify click activities as legitimate or fraudulent. In addition, a Deep Neural Network (DNN) model is employed to capture complex patterns and hidden relationships within large-scale advertising datasets. The hybrid approach combines the strengths of traditional machine learning and deep learning methods to improve detection accuracy and reduce false positives. Experimental results demonstrate that the proposed model effectively identifies fraudulent ad clicks and enhances the reliability of digital advertising systems. The system can assist advertising platforms and marketers in minimizing financial losses, improving campaign transparency, and ensuring fair advertising practices. This research contributes to the development of intelligent and scalable fraud detection mechanisms for modern online advertising