Abstract :

Modern Digital Infrastructures Including Enterprise Networks, Cloud Computing Platforms, IoT Ecosystems, And Online Service Environments Continuously Generate Vast Volumes Of Real-time Data Such As Network Packets, Authentication Logs, And System Performance Metrics. This Rapid Data Generation Necessitates Efficient, Intelligent, And Real-time Security Monitoring Mechanisms. Conventional Security Approaches Rely On Manual Log Analysis, Rule-based Intrusion Detection Systems, And Static Threshold Techniques, Which Are Time-consuming, Labor-intensive, And Unable To Adapt To Evolving Cyber Threats. As A Result, They Often Lead To Delayed Responses, High False Positive Rates, Limited Scalability, And Increased Vulnerability To Sophisticated Attacks. To Address These Challenges, This Work Proposes An Automated And Scalable AI-driven Security Framework For Real-time Anomaly Detection And Authentication Threat Analysis. Initially, Machine Learning Models Such As K-Nearest Neighbor (KNN) And Support Vector Classifier (SVC) Are Employed To Learn Network Behavior Patterns And Distinguish Between Normal And Malicious Activities. However, These Models Exhibit Limitations Including High Computational Complexity, Sensitivity To Feature Scaling, And Inefficiency With Large-scale Or Probabilistic Data. To Overcome These Issues, A Naive Bayes Classifier (NBC) Is Adopted As The Primary Probabilistic Inference Model. By Leveraging Bayesian Decision Theory And Modeling Conditional Feature Dependencies, NBC Enables Efficient Threat Probability Estimation With Reduced Computational Overhead And Improved Scalability For High-dimensional Datasets. The System Integrates Data Preprocessing, Class Balancing, Multi-model Training, And Deployment Through A Flask-based Web Interface. Performance Evaluation Using Accuracy, Precision, Recall, And F1-score Demonstrates Reliable Anomaly Detection, Validating The Framework As A Robust And Efficient Solution For Real-time Security Intelligence.

Published:

10-4-2026

Issue:

Vol. 26 No. 4 (2026)

Page Nos:

2191-2198

Section:

Articles

License:

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

How to Cite