Published Paper Details:

Malware is a major problem for everybody using the internet today. Malware that can quickly change its behaviour in response to new environments or security measures is called polymorphic malware. In order to evade detection by signature-based malware detection methods, polymorphic malware is continually changing its characteristic qualities. We employed several machine learning strategies for detecting malware and other dangerous threats. The high detection ratio is proof that the system employs the most effective algorithm possible. Accuracy information was enhanced by the confusion matrix's measurement of false positives and false negatives. Modern malware is highly developed and intricately designed to cause maximum damage. Most sophisticated malware is resilient and can evade detection for long periods of time. In this research, we'll be looking at a particularly sophisticated form of malware called advanced persistent threats (APTs). It is possible to attribute most cases of cyber espionage and sabotage to APTs. Complex, custom-tailored, and stealthy up until the point of compromise, advanced persistent threats (APTs) are difficult to detect. Automated, target-specific malware is deployed within a host or network by APTs so that assaults can be launched on demand in response to constant monitoring. In this paper, we will summarise and evaluate the state-of-the-art approaches to detection from several fields of study.