[ML101] Machine Learning Categories

Unsupervised Learning

Unsupervised Learning is when output variables are unlabeled.

Instead of looking into output prediction, unsupervised learning helps us uncover hidden patterns and relationships between input variables. Unsupervised learning work great with clustering analysis. By identifying subtle correlations between data, grouping, clustering can be done. Trends can be discovered.

Supervised Learning

independent variables X to dependent variable Y , look at y = f(x) For example, price of a house (Y, dependent variable) is dependent on many other independent variables X like, size, location so on. To train a supervised model, both input and out are needed, whihc known as labeled. Examples of algorithms used for supervised learning include regression analysis, decison trees, k-nearest neighbor and svm.