Glossary - fmxy/k-means GitHub Wiki

This page defines terms and phrases used throughout the documentation of the project (javadoc, comments, commit messages etc.)

• (cluster) centroid = single value to express the contents of a cluster; the cluster centroids in kmeans are created by calculating the mean of all elements in the cluster; sometimes referred to as: (cluster) mean/center

• run parameter/argument = command line argument to be provided to the algorithm in order to run a specific strategy or adapt amount of clusters/data points/ iterations

• strategy = with respect to the well-established strategy pattern a run strategy in this project means a specific way of running the classic kmeans clustering algorithm (while not changing the algorithm steps itself). E.g. the algorithm can be run with a 'sequential' or 'thread-based' strategy.

• (data) point = element of a data set that is to be clustered. Holds positive values between 0 and 1. Data points can be read from a file or randomly generated in this project. The prior is the default to provide reproducible results.