Machine Learning
Solving problems without providing explicit instructions for all parts of the solution.
Reinforcement Learning
Wikipedia machine learning reinforcement learning
You provide rules and a way to judge a solution. The reinforcement learning will than find better and better solutions on its own. One use case are AI players for games.
Unsupervised Learning
Find solutions without you needing to providing training data
Clustering
Cluster data points together, see also DBSCAN
Dimensional Reduction
Wikipedia dimensional reduction
Transfer an n dimensional data set into a m dimensional one, where m<n