Tashkent State University of Oriental Studies International Scientific-Practical Conference

NAME: Sumanshu Sohal

Designation: FireEye SOC Analyst, Washington DC, USA

Country: USA

SPEAKER ID: ISPC-7

CONFERENCE DATES: 20-22 SEP 2025 (Historical Realities and Philosophical Ideals: Dialogues Between Disciplines) Tashkent State University of Oriental Studies International Scientific-Practical Conference.

CONFERENCE TYPE: Hybrid

SECTION: Artificial Intelligence

CONFERENCE SPEECH TITLE : A Comparative Analysis of Supervised vs. Unsupervised Machine Learning Algorithms for False Positive Reduction in Hybrid Cloud SIEM Environments

SHORT DESCRIPTION : This talk explores the challenge of false positive alerts in hybrid cloud Security Information and Event Management (SIEM) systems and the role of machine learning in addressing them. Using the CIC-IDS2017 dataset, I present a comparative analysis of supervised algorithms (Random Forest, XGBoost) and unsupervised approaches (Isolation Forest, Autoencoders). The findings highlight that supervised models excel at reducing known false positives with high accuracy, while unsupervised methods are more effective at capturing unusual but benign patterns. Attendees will gain actionable insights into selecting and deploying intelligent ML-driven strategies that enhance detection precision, reduce analyst fatigue, and improve operational efficiency in modern security operations centers.