This thesis examines using machine learning methods to extract cyber threat intelligence from hacker forums. It proposes a two-phase process using supervised and unsupervised learning. In phase one, classifiers like support vector machines are used to classify forum posts as relevant or not to security. Phase two applies topic modeling to identified relevant posts to discover discussion themes. Experiments on a real hacker forum show these methods effectively identify security-related information including zero-days, credentials, and malware. The study demonstrates hacker forums can provide useful threat intelligence and machine learning helps analyze large amounts of forum data.