## About

Unsupervised learning is the second type of function that an algorithm can perform.

The algorithm is said to be unsupervised when no response is used in the algorithm.

Unsupervised Learning has the goal of discovering relationships and patterns rather than of determining a particular value as in supervised learning. There is:

- no outcome variable, just a set of predictors (features). It is non-directed.
- no distinction between dependent and independent attributes.
- no previously-known result to guide the algorithm in building the model.

Unsupervised learning can be used:

- for descriptive purposes.
- to find groups of samples that behave similarly,
- to find features that behave similarly,
- to find linear combinations of features with the most variation. (pca ?)
- to classify (Clustering models can be applied to classify cases according to their cluster assignments)

Unsupervised learning can be useful as a pre-processing step of supervised learning.

## Articles Related

## Algorithm

## Example

Example | Illustrate the Model |
---|---|

Determine distinct segments of a population and the attribute values indicating an individual’s membership in a particular segment | Clustering |

Determine the five items most likely to be purchased at the same time as item X. | Market Basket Analysis (or association) |