A kurtosis-based dynamic approach to Gaussian mixture modeling
| dc.contributor.author | Vlassis, N. | en |
| dc.contributor.author | Likas, A. | en |
| dc.date.accessioned | 2015-11-24T17:03:13Z | |
| dc.date.available | 2015-11-24T17:03:13Z | |
| dc.identifier.issn | 1083-4427 | - |
| dc.identifier.uri | https://olympias.lib.uoi.gr/jspui/handle/123456789/11143 | |
| dc.rights | Default Licence | - |
| dc.subject | expectation-maximization (em) algorithm | en |
| dc.subject | gaussian mixture modeling | en |
| dc.subject | number of mixing kernels | en |
| dc.subject | probability density function estimation | en |
| dc.subject | total kurtosis | en |
| dc.subject | weighted kurtosis | en |
| dc.subject | probabilistic neural networks | en |
| dc.subject | maximum-likelihood | en |
| dc.subject | em algorithm | en |
| dc.subject | components | en |
| dc.subject | number | en |
| dc.title | A kurtosis-based dynamic approach to Gaussian mixture modeling | en |
| heal.abstract | We address the problem of probability density function estimation using a Gaussian mixture model updated with the expectation-maximization (EM) algorithm. To deal with the case of an unknown number of mixing kernels, we define a new measure for Gaussian mixtures, called total kurtosis, which is based on the weighted sample kurtoses of the kernels. This measure provides an indication of how well the Gaussian mixture fits the data. Then we propose a new dynamic algorithm for Gaussian mixture density estimation which monitors the total kurtosis at each step of the Ehl algorithm in order to decide dynamically on the correct number of kernels and possibly escape from local maxima. We show the potential of our technique in approximating unknown densities through a series of examples with several density estimation problems. | en |
| heal.access | campus | - |
| heal.fullTextAvailability | TRUE | - |
| heal.journalName | Ieee Transactions on Systems Man and Cybernetics Part a-Systems and Humans | en |
| heal.journalType | peer reviewed | - |
| heal.language | en | - |
| heal.publicationDate | 1999 | - |
| heal.recordProvider | Πανεπιστήμιο Ιωαννίνων. Σχολή Θετικών Επιστημών. Τμήμα Μηχανικών Ηλεκτρονικών Υπολογιστών και Πληροφορικής | el |
| heal.type | journalArticle | - |
| heal.type.el | Άρθρο Περιοδικού | el |
| heal.type.en | Journal article | en |
Αρχεία
Φάκελος/Πακέτο αδειών
1 - 1 of 1
Φόρτωση...
- Ονομα:
- license.txt
- Μέγεθος:
- 1.74 KB
- Μορφότυπο:
- Item-specific license agreed upon to submission
- Περιγραφή: