<p>In view of the evolving domain of cyber threats, traditional signature based detection tools encounter difficulties in detecting malware. The conventional approach utilizing pattern matching across database is easily undermine by newer threats. Machine learning is able to intelligently detect and classify malware based on the attack pattern which serves as an alternative approach to the enhanced anti-malware solution. This research therefore presents a comprehensive study in the elucidation on the performance of various machine learning models for real-time malware detection and classification. The work is extended to cover three main scope of studies: malware detection, malware classification, and network packet screening. The malware detection component uses static analysis approaches to distinguish between malicious and legitimate files based on the information of the portable executable headers. After detection, the malware classification component can be used to categorize the identified malware into specific malware families, providing deeper insights into the malware characteristics. Additionally, the network packet classification component analyzes network traffic, identifying packets as benign or malicious. In addition, a dedicated application is constructed for automated feature extraction and result visualization in real-time. Experimental outcome demonstrates promising results of the machine learning enabled anti-malware solution for real-time employment, underscoring its prevalence for effective prevention of related cyber threats.</p>