DEVELOPMENT OF A MALWARE DETECTION SYSTEM TO IDENTIFY AND CATEGORIZE MALICIOUS SOFTWARE

The emerging dangers of malware growth pose new threats to data security, user privacy, data integrity, and the reliability of personal, corporate, and government networks. This project designed and implemented a malware detection system to detect and categorize the software threats as malware. The objectives included understanding malware trends and attack patterns, designing a detection system using machine learning, implementing the designed system, and testing and evaluating the designed system. The need for intelligent, adaptive, and easy-to-use detection systems was highlighted by the shortcomings of conventional signature-based antivirus software against new malware threats. The methodology included study of pertinent academic papers, cybersecurity reports, and case studies to understand the malware behaviour and attack trends. System design was developed using flowcharts and UML modelling tools to capture system architecture. The web interface was developed using HTML and PHP, and the implementation of the machine learning model was carried out using Python. Model training and testing were conducted using Scikit-learn and TensorFlow, with feature extraction based on executable file attributes. The system was tested by multiple users, and its performance was measured using accuracy, precision, and recall as evaluation metrics. The system was found to achieve a high detection accuracy of 0.99, precision of 0.99, and recall of 0.98 in detecting both known and unknown threats, proving its capability for both known and unknown threat detection. Future work should focus on integrating cloud-based updates and expanding the dataset to strengthen defense against emerging malware threats. Keywords: Malware detection, Machine learning, Classification model, Threat analysis, Web interface Word Count: 243