论文信息 - A New Methodology to Compare Clustering Algorithms

A New Methodology to Compare Clustering Algorithms

In the context of unsupervised clustering, lots of different algorithms have been proposed. Most of them consist in optimizing an objective function using a search strategy. We present here a new methodology for studying and comparing the performances of the objective functions and search strategies employed.

Céline Robardet | Fabien Feschet

[1] Hugo Steinhaus,et al. On a certain distance of sets and the corresponding distance of functions , 1958 .

[2] M. Karonski,et al. On Marczewski-Steinhaus type distance between hypergraphs , 1977 .

[3] Anil K. Jain,et al. Algorithms for Clustering Data , 1988 .

[4] Douglas H. Fisher,et al. Iterative Optimization and Simplification of Hierarchical Clusterings , 1996, J. Artif. Intell. Res..

[5] Jack Edmonds,et al. Maximum matching and a polyhedron with 0,1-vertices , 1965 .