Impact Factor

7.883 Call For Paper

Share on

Author(s):

Giritharan R | Dr. K. Annalakshmi
Keywords:

Cybersecurity, Machine Learning, Intrusion Detection System, Random Forest, AdaBoost, Bernoulli Naive Bayes, Network Threat Detection, Django, Real-Time Monitoring.
Abstract:
The Rapid Growth Of Digital Technologies Has Substantially Increased Exposure To Cybersecurity Threats Including Malware, Phishing, Ransomware, And Unauthorized Network Intrusions. Traditional Signature-based Security Systems Are Inherently Reactive And Fail Against Zero-day Exploits And Polymorphic Attacks. This Paper Presents ThreatGuardian, An AI-powered Cybersecurity Threat Detection And Prevention Framework That Leverages Machine Learning Algorithms — Random Forest (RF), AdaBoost Classifier (ADC), And Bernoulli Naive Bayes Classifier (BNC) — To Identify And Classify Malicious Network Activities In Real Time. The System Integrates Data Preprocessing, Exploratory Data Analysis, Feature Extraction, And Model Evaluation Pipelines With A Django-based Web Interface For Practical Deployment. Trained And Evaluated On A Publicly Available Network Intrusion Dataset From Kaggle, The Best-performing Model Is Serialized And Deployed For Real-time Inference. Performance Evaluation Using Accuracy, Precision, Recall, And F1-score Demonstrates That The Proposed Framework Significantly Outperforms Traditional Rule-based Methods, Providing An Adaptive, Scalable, And User-accessible Solution For Modern Cyber Threat Management.