As malware increases nowadays, it is necessary to safeguard your system from the malware. Malware is being protected by traditional methods but it only protects system from the malware whose signature is known. So we aim to prepare a software which detects the malware (signature is unknown) by the malware(signature is known) and after detecting the malware, steps must be taken to retrain the model and disallow the malware to compromise the system. For detecting the malware already known machine learning methods such as random forest classifier, multi view ensemble learning etc. will be used.