ABSTRACT

Cyber vulnerabilities pose a threat across systems in the Department of Defense. Finding ways to analyze network traffic and detect malicious behavior on a network will help keep these systems safe. This paper looks at the data collection techniques, model creation, and results of building a recurrent neural network to classify incoming traffic as normal or malicious. Additionally, it considers how the information will be best portrayed on a GUI to network administrators. The model’s initial accuracy is 83.45% when trained on 500,017 connections. With increased accuracy, this tool may be used by the Department of Defense to help defend its networks.

BIO

Raised in Darien, Connecticut, Madeleine Schneider is a double major in computer science and mathematical sciences at the United States Military Academy at West Point. Outside of the classroom Madeleine participated in three years of Division I Track and Field and held multiple leadership roles within her academic company. During the summer of 2017 Madeleine worked at Lawrence Livermore National Labs with the Cyber Defenders program and during the summer of 2018 she conducted research with the National Security Agency. Madeleine is continuing her research focus this year with three separate research projects to include network security with machine learning, statistical analysis of machine learning bias, and parallel computing in the analysis of big data. Madeleine hopes to continue pursuing her research interests in machine learning to better prepare herself for a career as an Army Cyber officer. Madeleine is a Marshall Scholar.