Person:
Lagulos, Marlon

Profile Picture
Email Address
Birth Date
Biography URL
https://www.husson.edu/directory/marlon_lagulos
Term at University
Research Projects
Organizational Units
Job Title
Instructor, College of Business
Last Name
Lagulos
First Name
Marlon
Name
Degrees Held
M.Eng. Engineering, University of Mindanao
B.S. Electronics and Communications Engineering, University of Southeastern Philippines

Filters

20261
Reset filters

Settings

Author: Lagulos, Marlon × End date: 2026 ×

Search Results

Now showing 1 - 1 of 1
  • Thumbnail Image
    Publication
    EARLY WARNING FOR CAMPUS CYBER RISKS
    (2026-04-16) Mwiseneza, Elvis; Lagulos, Marlon
    Cyberattacks are a growing problem for colleges and universities. A single phishing email, stolen password, or infected device can interrupt classes, expose private records, and create major costs for colleges and universities. This project presents XCampus cyber risk monitor, a small lightweight cybersecurity risk monitor designed to give early warning signs before a major or minor incident happens. Instead of only detecting attacks after they begin, the system looks for behavior patterns by learning the pattern and that shows risk is increasing over time. It analyzes six main types of activity such as login failures, suspicious IP access, unusual file downloads, new device connections, phishing email patterns, and privileged access attempts from users. Because real campus security data is private, the project uses realistic designed synthetic data to simulate activity across departments such as finance, admissions, registrar, library, research, and student services. The data is cleaned and grouped by department and system, then transformed into features such as off-hours activity, sudden event spikes, and changes in external access. These features are used in three approaches which are; a machine learning model for risk prediction, an anomaly detection model for unusual behavior, and a rule based baseline for comparison. The system combines these outputs into a final risk score and classifies each department or system as low, medium, or high risk. Results from the prototype, it shows that the predictive model performed better than the rule based method, while anomaly detection helped to show suspicious patterns. The final product includes a simple dashboard for viewing risk levels and trends. This study shows that a practical and low cost early warning system can help universities move from reactive security toward proactive cyber risk management.