Date of Award
2017
Type
Thesis
Major
Computer Science - Applied Computing Track
Department
TSYS School of Computer Science
Abstract
Anomaly detection in user access patterns using artificial neural networks is a novel way of combating the ever-present concern of computer network intrusion detection for many entities around the world. Anomaly detection is a technique in network security in which a profile is built around a user's normal daily actions. The data collected for these profiles can be as following: file access attempts; failed login attempts; file creations; file access failures; and countless others. This data is collected and used as training data for a neural network. There are many types of neural networks, such as multi-layer feed-forward network; recurrent networks; support vector machines; and others. For our study, we implemented our own self¬ organizing map (SOM), which we found to not be as heavily researched as other neural network approaches. Using the KDD Cup 99 dataset, we compared our own SOM implementation against other neural network implementations and determine the effectiveness of such an approach.
Recommended Citation
Parrachavez, Manuel R., "Using Self-Organizing Maps for Computer Network Intrusion Detection" (2017). Theses and Dissertations. 293.
https://csuepress.columbusstate.edu/theses_dissertations/293
