This is a paper of the MCP project of class EAP-508-P02. In this project, we select one topic in our major filed, and do a literature review about this topic. My topic is K-Means variants for web-scale clustering.

Tremendous high-dimension data of web-scale clustering make K-Means, one popular clustering solution(Jain,2009), becomes unbearably slow(Zhao et al.,2016). Applications requiring instant results even optimized K-Means can not meet facing web-scale data(Sculley,2010). Many web-scale K-Means variants have been proposed (Sculley,2010;Broder et.al.,2014;Zhao et.al,2010), but no article review these methods. We review web-scale K-Means variants to summarize and compare them. Accommodating K-Means algorithm to web-scale clustering is a new research trend, this project offers reference to future researches and help novices have a clearer mind about K-Means variants for web-scale clustering.

Download: comparing-k-means-variants-for-web-scale-clustering