Culminating Project Title
Date of Award
Culminating Project Type
Information Assurance: M.S.
Information Assurance and Information Systems
Herberger School of Business
Creative Commons License
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License.
Keywords and Subject Headings
Machine Learning, Deep Learning, Database Auditing, Artificial Neural Networks, A. Daghighi
Data auditing is a fundamental challenge for organizations who deal with large databases. Databases are frequently targeted by attacks that grow in quantity and sophistication every day, and one-third of which are coming from users inside the organizations. Database auditing plays a vital role in protecting against these attacks. Native features in data base auditing systems monitor and capture activities and incidents that occur within a database and notify the database administrator. However, the cost of administration and performance overhead in the software must be considered. As opposed to using native auditing tools, the better solution for having a more secure database is to utilize third-party products. The primary goal of this thesis is to utilize an efficient and optimized deep learning approach to detect suspicious behaviors within a database by calculating the amount of risk that each user poses for the system. This will be accomplished by using an Artificial Neural Network as an enhanced feature of analyzer component of a database auditing system. This ANN will work as a third-party product for the database auditing system. The model has been validated in order to have a low bias and low variance. Moreover, parameter tuning technique has been utilized to find the best parameters that would result in the highest accuracy for the model.
Daghighi, Amirali, "Application of an Artificial Neural Network as a Third-Party Database Auditing System" (2019). Culminating Projects in Information Assurance. 86.