Detecting clusters in moderate-to-high dimensional data: subspace clustering, pattern-based clustering, and correlation clustering

摘要: As a prolific research area in data mining, subspace clustering and related problems induced vast amount of proposed solutions. However, many publications compare new proposition -- if at all with one or two competitors even so called "naive" ad hoc solution but fail to clarify the exact problem definition. consequence, solutions are thoroughly compared experimentally, it will often remain unclear whether both tackle same or, they do, agree certain tacit assumptions how such may influence outcome an algorithm. In this tutorial, we try (i) different definitions general, (ii) specific difficulties encountered field research, (iii) varying assumptions, heuristics, intuitions forming basis approaches, (iv) several prominent essentially problems.