Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{193094,
        author = {K. Veerendra kumar and Golla veeresh and Kandlagari santhosh reddy and Kuruva siva and Mrs. S.S. RAJA KUMARI and Dr. P. Veeresh},
        title = {A Improving Ad Click Fraud Detection Through Machine Learning and Deep Learning Models},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {12},
        number = {9},
        pages = {3654-3660},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=193094},
        abstract = {Due to the swift development of the mobile advertising market, the cases of fraudulent clicks of advertisements have grown considerably, causing considerable financial expenses to advertisers. This work provides a comparative research on different machine learning (ML) and deep learning (DL) models to identify ad click fraud. The suggested solution would combine classical machine learning (Logistic Regression, Random Forest, and XGBoost) and deep learning (Convolutional Neural Networks, CNN, Long Ssshort-Term Memory, and Gated Recurrent Units, respectively) in a single solution. Such models are chosen on ground of their ability to differentiate legit and fraudulent ad-click. In order to further boost the performance of the detection, Stacking Classifier is used to utilize the positive aspects of several models. The implementation of the system is based on Flask to facilitate fraud detection in mobile advertising campaigns and offer an easy way of tracking and minimizing the losses. The results of the experiments also prove that the Maximum accuracy of 92 was achieved by the Random Forest, and the deep learning models including CNN and LSTM demonstrated competitive results of 90 and 91, respectively. The Stacking Classifier was found to be more effective with a balanced precision-recall score of 0.92, which shows that the classifier gives good results in fraudulent clicks. The proposed system will contribute to increasing the validity of mobile advertising platforms and minimizing the financial loss caused by ad click fraud recognition.},
        keywords = {Ad Click Fraud, Machine Learning, Deep Learning, Fraud Detection, Stacking Classifier, Random Forest, XGBoost, Logistic Regression, CNN, LSTM.},
        month = {February},
        }