Abstract | ||
---|---|---|
Previous work in mixture model clusteringhas focused primarily on the issue of modelselection. Model scoring functions (includingpenalized likelihood and Bayesian approximations)can guide a search of the model parameterand structure space. Relatively littleresearch has addressed the issue of howto move through this space. Local optimizationtechniques, such as expectation maximization,solve only part of the problem; westill need to move between different local optima.The... |
Year | Venue | Keywords |
---|---|---|
2001 | ICML | repairing faulty mixture,density estimation,expectation maximization,score function,mixture model |
Field | DocType | ISBN |
Density estimation,Pattern recognition,Computer science,Artificial intelligence,Mixture model | Conference | 1-55860-778-1 |
Citations | PageRank | References |
12 | 2.43 | 7 |
Authors | ||
2 |
Name | Order | Citations | PageRank |
---|---|---|---|
Peter Sand | 1 | 351 | 22.63 |
Andrew W. Moore | 2 | 113 | 17.17 |