Please use this identifier to cite or link to this item:
http://hdl.handle.net/20.500.12188/24478
DC Field | Value | Language |
---|---|---|
dc.contributor.author | Madevska Bogdanova, Ana | en_US |
dc.contributor.author | Nikolikj, D | en_US |
dc.date.accessioned | 2022-11-21T08:21:06Z | - |
dc.date.available | 2022-11-21T08:21:06Z | - |
dc.date.issued | 2002 | - |
dc.identifier.uri | http://hdl.handle.net/20.500.12188/24478 | - |
dc.description.abstract | We present an alternative way of interpreting and modifying the outputs of the Support Vector Machine (SVM) classifiers – method MSVMO (Modified SVM Outputs). Stemming from the geometrical interpretation of the SVM outputs as a distance of individual patterns from the hyperplane, allows us to calculate its posterior probability i.e. to construct a probabilitybased measure of belonging to one of the classes, depending on the vector’s relative distance from the hyperplane. We illustrate the results by providing suitable analysis of three classification problems and comparing them with an already published method for modifying SVM outputs. | en_US |
dc.publisher | Institute of Informatics, Faculty of Natural Sciences and Mathematics, Ss. Cyril and Methodius University in Skopje, Macedonia | en_US |
dc.subject | Support Vector Machines; pattern classification; modified outputs; post-processing; posterior probability | en_US |
dc.title | Svm Classifiers with Moderated Outputs for Automatic Classification in Molecular Biology | en_US |
dc.type | Proceedings | en_US |
dc.relation.conference | Third International Conference on Informatics and Information Technology | en_US |
item.fulltext | With Fulltext | - |
item.grantfulltext | open | - |
Appears in Collections: | Faculty of Computer Science and Engineering: Conference papers |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
3ndCiiT-10.pdf | 431.5 kB | Adobe PDF | View/Open |
Page view(s)
35
checked on Jul 22, 2024
Download(s)
8
checked on Jul 22, 2024
Google ScholarTM
Check
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.