A fast version of the k-means classification algorithm for astronomical applications

Abstract

Context. K-means is a clustering algorithm that has been used to classify large datasets in astronomical databases. It is an unsupervised method, able to cope very different types of problems. Aims. We check whether a variant of the algorithm called single-pass k-means can be used as a fast alternative to the traditional k-means. Methods. The execution time of the two algorithms are compared when classifying subsets drawn from the SDSS-DR7 catalog of galaxy spectra. Results. Single-pass k-means turn out to be between 20 % and 40 % faster than k-means and provide statistically equivalent classifications. This conclusion can be scaled up to other larger databases because the execution time of both algorithms increases linearly with the number of objects. Conclusions. Single-pass k-means can be safely used as a fast alternative to k-means.

References in corpus (11)

• The Seventh Data Release of the Sloan Digital Sky Survey
• Finding Anomalous Periodic Time Series: An Application to Catalogs of Periodic Variable Stars
• Systematic search for extremely metal poor galaxies in the Sloan Digital Sky Survey
• Automatic unsupervised classification of all SDSS/DR7 galaxy spectra
• Automated unsupervised classification of the Sloan Digital Sky Survey stellar spectra using k-means clustering
• Qualitative interpretation of galaxy spectra
• A six-parameter space to describe galaxy diversification
• Search for blue compact dwarf galaxies during quiescence II: metallicities of gas and stars, ages, and star-formation rates
• Do galaxies form a spectroscopic sequence?
• Uncovering the formation of ultra-compact dwarf galaxies by multivariate statistical analysis
• Spectral matching for abundances and clustering analysis of stars on the giant branches of ω Centauri

Cited by in corpus (4)

• Search for Extremely Metal-poor Galaxies in the Sloan Digital Sky Survey (II): high electron temperature objects
• Data mining techniques on astronomical spectra data. I : Clustering Analysis
• A Systematic Study Of Superluminous Supernova Lightcurve Models Using Clustering
• Modeling quasar variability through self-organizing map-based process