Show simple item record

Measure of Similarity between GMMs by Embedding of the Parameter Space That Preserves KL Divergence

dc.contributor.authorBranislav Popović
dc.contributor.authorLenka Cepova
dc.contributor.authorRobert Cep
dc.contributor.authorMarko Janev
dc.contributor.authorLidija Krstanović
dc.contributor.otherFaculty of Technical Sciences, University of Novi Sad, Trg D. Obradovića 6, 21000 Novi Sad, Serbia
dc.contributor.otherDepartment of Machining, Faculty of Mechanical Engineering, VSB-Technical University of Ostrava, Assembly and Engineering Metrology, 17. listopadu 2172/15, 708 00 Ostrava Poruba, Czech Republic
dc.contributor.otherDepartment of Machining, Faculty of Mechanical Engineering, VSB-Technical University of Ostrava, Assembly and Engineering Metrology, 17. listopadu 2172/15, 708 00 Ostrava Poruba, Czech Republic
dc.contributor.otherInstitute of Mathematics, Serbian Academy of Sciences and Arts, Kneza Mihaila 36, 11000 Belgrade, Serbia
dc.contributor.otherFaculty of Technical Sciences, University of Novi Sad, Trg D. Obradovića 6, 21000 Novi Sad, Serbia
dc.date.accessioned2021-04-26T00:01:25Z
dc.date.available2025-10-02T03:33:32Z
dc.date.issued01-04-2021
dc.identifier.issn-
dc.identifier.urihttps://www.mdpi.com/2227-7390/9/9/957
dc.description.abstractIn this work, we deliver a novel measure of similarity between Gaussian mixture models (GMMs) by neighborhood preserving embedding (NPE) of the parameter space, that projects components of GMMs, which by our assumption lie close to lower dimensional manifold. By doing so, we obtain a transformation from the original high-dimensional parameter space, into a much lower-dimensional resulting parameter space. Therefore, resolving the distance between two GMMs is reduced to (taking the account of the corresponding weights) calculating the distance between sets of lower-dimensional Euclidean vectors. Much better trade-off between the recognition accuracy and the computational complexity is achieved in comparison to measures utilizing distances between Gaussian components evaluated in the original parameter space. The proposed measure is much more efficient in machine learning tasks that operate on large data sets, as in such tasks, the required number of overall Gaussian components is always large. Artificial, as well as real-world experiments are conducted, showing much better trade-off between recognition accuracy and computational complexity of the proposed measure, in comparison to all baseline measures of similarity between GMMs tested in this paper.
dc.format-
dc.language.isoEN
dc.publisherMDPI AG
dc.relation.uri['https://iteecs.com/index.php/iteecs/about#focus-scope', 'https://iteecs.com/index.php/iteecs/index', 'https://iteecs.com/index.php/iteecs/submission#author-guidelines', 'https://iteecs.com/index.php/iteecs/submission#publication-charges']
dc.rightsCC BY
dc.subject['electrical engineering', 'computer engineering', 'electronics engineering', 'artificial intelligence', 'green energy', 'control engineering', 'Electrical engineering. Electronics. Nuclear engineering', 'TK1-9971']
dc.subject.lccMathematics
dc.titleMeasure of Similarity between GMMs by Embedding of the Parameter Space That Preserves KL Divergence
dc.typeArticle
dc.description.keywordsGaussian mixture models
dc.description.keywordssimilarity measures
dc.description.keywordsdimensionality reduction
dc.description.keywordsKL-divergence
dc.description.pages-
dc.description.doi10.3390/math9090957
dc.title.journalMathematics
dc.identifier.e-issn2227-7390
dc.identifier.oaioai:doaj.org/journal:b8d5163330fc42fda1c0955150498e18
dc.journal.infoVolume 9, Issue 9


This item appears in the following Collection(s)

Show simple item record