Cyber Security researchers, professionals and enthusiasts are currently faced with the challenge of coming up with advanced methods to deal with the bludgeoning of highly sophisticated Malware attacks on critical infrastructures housing sensitive information. The task of detecting, analysing and investigating the presence of Malware in a system is a daunting one, necessitating deep knowledge and skills in Malware Analysis. This analysis helps practitioners to uncover the behaviour and also the characteristics of Malware, thereby enabling Cyber Security practitioners and businesses to be proactive in building tools and techniques for defence against this type of Cyber-attacks. Malware authors are becoming more sophisticated in the methods they are using to evade detection. State of the art techniques are being applied to Malware Analysis with Machine Learning increasingly becoming popular while yielding very promising results. This paper proposes a hybrid framework for Unsupervised and Supervised Machine Learning methods for dynamic and static Malware analysis.