Publication Details
SVM CLASSIFIERS CREATION IN PARALLEL CONSTRAINED ENVIRONMENT
Zemčík Pavel, prof. Dr. Ing. (DCGM FIT BUT)
Herout Adam, prof. Ing., Ph.D. (DCGM FIT BUT)
Beran Vítězslav, doc. Ing., Ph.D. (DCGM FIT BUT)
Support Vector Machine, SVM, Sun Grid Engine, dataset, Feature vectors, Parametric training
Support Vector Machines (SVM) classification is one of the most frequently used classification methods based on
machine learning used today. SVMs, however, are dependent on many parameters and settings and so it is suitable to
perform the learning process in many instances and evaluate what parameters and settings are suitable for each individual
case of data and task. This paper focuses on a novel framework that allows parametric training of SVM classifiers in
parallel computer environment which has certain constraints regarding the resources available to the training task and
duration of it. The framework is introduced and conclusions are drawn.
This paper focuses on a novel framework that allows parametric training of SVM classifiers in parallel computer environment which has certain constraints regarding the resources available to the training task and duration of it.
@INPROCEEDINGS{FITPUB9273, author = "Ivo \v{R}ezn\'{i}\v{c}ek and Pavel Zem\v{c}\'{i}k and Adam Herout and V\'{i}t\v{e}zslav Beran", title = "SVM CLASSIFIERS CREATION IN PARALLEL CONSTRAINED ENVIRONMENT", pages = "535--538", booktitle = "Proc. of the IADIS Int. Conf. - Computer Graphics, Visualization, Computer Vision and Image Processing, CGVCVIP 2010, Visual Commun., VC 2010, Web3DW 2010, Part of the MCCSIS 2010", year = 2010, location = "Freiburg im Breissgau, DE", publisher = "IADIS", ISBN = "978-972-8939-22-9", language = "english", url = "https://www.fit.vut.cz/research/publication/9273" }