Machine learning

From stats++ wiki
Jump to: navigation, search

Broadly speaking, machine learning methods aim to extract knowledge from data, which is (often) not a priori obvious. Such data can be either unlabeled or labeled; both consist of a set input {x0, x1, ..., xn}, but labeled contains a set of corresponding output {y0, y1, ..., xn}. For the former data type (unlabeled), ML methods attempt to learn inherent patterns in the (input) data; they can also do the same for the latter type (labeled), but also learn the functional relationship between the input and output (regresssion).

stats++ includes a number of machine learning algorithms.

Supervised learning

Unsupervised learning