With the advent and subsequent explosion of the internet, global connectivity has been achieved, and is on the rise. This provides a host of advantages such as connectivity and communication, information broadcast and transmission, amongst others. This however introduces a new set of challenges: the safety and protection of these communication channels amongst them. Information has always been power, and the widespread mature of information only results in the widespread attempts to procure it, sometimes via illegal channels. In view of this, this research aims at detecting Crypto-ransomware and locker ransomware. Data was collected from an open repository and cleaned. The cleaned data was then split into tests, train sets and validation which was used to train a number of ML models based on the: Random Forest algorithm, Support Vector Machine (SVM) and Gradient boosting algorithm. Ransomware is one of the well-known ways and frequent use which cyber-attackers use in infecting their victims, either through phishing or drive download. Attackers will create an email pretending to be from a genuine resource and send it to their targeted victims. However, this research illustrated how to combat crypto-ransomware and locker ransomware. Implementing the machine learning algorithm, the system can detect ransomware under 30’s, giving computer users over 90% assurance of their system for ransomware free.