Time-constrained Detection of Network Attacks


Yeeun Kim

Document Type


Degree Name

Master of Science (MS)


Computer Science

Date of Award

Spring 2022


Time-constrained detection of network attacks is very important in modern society to protect networks exposed to various cyber-attack risks. Anomaly detection is also an important task, but events can be limited by various environmental conditions. For Time-constrained detection of network attacks, the main challenges are (1) time-constrained data processing, (2) methodology for time-constrained network attack identification, and (3) Improved model performance among various variables. To deal with these challenges, this research will explore time-constrained identification of attacks with the following three angles: (1) Adding problem dataset preprocessing; (2) Deep learning models including Temporal Convolutional Network (TCN), Bayesian Temporal Convolutional Network (BTCN); (3) Find machine learning tuning. Based on the analysis of the above experiments’ results, this study will explore a novel model using machine learning, taking advantage of weakly-supervised learning by employing a few attacks labeled data for early attack identification.


Dongeun Lee

Subject Categories

Computer Sciences | Physical Sciences and Mathematics