Abstract | ||
---|---|---|
. We propose a new optimisation method for estimating boththe parameters and the structure, i. e. the number of components, of afinite mixture model for density estimation. We employ a hybrid methodconsisting of an evolutionary algorithm for structure optimisation in conjunctionwith a gradient-based method for evaluating each candidatemodel architecture. For structure modification we propose specific, problemdependent evolutionary operators. The introduction of a regularisationterm... |
Year | DOI | Venue |
---|---|---|
1998 | 10.1007/BFb0056941 | PPSN |
Keywords | Field | DocType |
density estimation models,evolutionary algorithms,evolutionary algorithm,density estimation,mixture model | Density estimation,Mathematical optimization,Evolutionary algorithm,Expectation–maximization algorithm,Computer science,Generalization,Model architecture,Evolutionary operators,Algorithm,Probability density function,Mixture model | Conference |
Volume | ISSN | ISBN |
1498 | 0302-9743 | 3-540-65078-4 |
Citations | PageRank | References |
3 | 0.50 | 6 |
Authors | ||
5 |
Name | Order | Citations | PageRank |
---|---|---|---|
Martin Kreutz | 1 | 40 | 5.27 |
Anja M. Reimetz | 2 | 3 | 0.50 |
Bernhard Sendhoff | 3 | 2272 | 240.31 |
Claus Weihs | 4 | 3 | 0.50 |
W von Seelen | 5 | 503 | 140.13 |