Exploring Effectiveness of HIDS for Machine-Learning Data Defense

Himani Vommi

As machine learning (ML) continues to support technological systems, the number of ways such systems can be compromised to cause unwanted behavior increases. Splunk, a cybersecurity monitoring tool, is also subject to advanced persistent threats (APTs) seeking to alter ML-powered security tools to evade detection. This research determines how effectively a HIDS (host-based intrusion detection system) can alert data integrity loss within the Splunk machine learning model to help security personnel detect an APT. 

Major: 
Cybersecurity Analytics and Operations
Exhibition Category: 
Engineering
Exhibition Format: 
Poster Presentation
Campus: 
Brandywine
Faculty Sponsor: 
John Landmesser
Poster Number: 
16525

Award Winner

Engineering - Second Place