ResearchSpace

Data characteristics that determine classifier performance

Show simple item record

dc.contributor.author Van der Walt, Christiaan M
dc.contributor.author Barnard, E
dc.date.accessioned 2007-07-26T07:46:44Z
dc.date.available 2007-07-26T07:46:44Z
dc.date.issued 2006-11
dc.identifier.citation Van der Walt, C and Barnard, E. 2006. Data characteristics that determine classifier performance. 17th Annual Symposium of the Pattern Recognition Association of South Africa, Parys, South Africa, 29 Nov - 1 Dec 2006, pp 6 en
dc.identifier.uri http://hdl.handle.net/10204/1038
dc.description This paper is published in the SAIEE Africa Research Journal, Vol 98(3), pp 87-93
dc.description.abstract The relationship between the distribution of data, on the one hand, and classifier performance, on the other, for non-parametric classifiers has been studied. It is shown that predictable factors such as the available amount of training data (relative to the dimensionality of the feature space), the spatial variability of the effective average distance between data samples, and the type and amount of noise in the data set influence such classifiers to a significant degree. The methods developed here can be used to gain a detailed understanding of classifier design and selection. en
dc.language.iso en en
dc.subject Data classifier performance en
dc.subject Datasets en
dc.subject Non-parametric classifiers en
dc.title Data characteristics that determine classifier performance en
dc.type Conference Presentation en
dc.identifier.apacitation Van der Walt, C. M., & Barnard, E. (2006). Data characteristics that determine classifier performance. http://hdl.handle.net/10204/1038 en_ZA
dc.identifier.chicagocitation Van der Walt, Christiaan M, and E Barnard. "Data characteristics that determine classifier performance." (2006): http://hdl.handle.net/10204/1038 en_ZA
dc.identifier.vancouvercitation Van der Walt CM, Barnard E, Data characteristics that determine classifier performance; 2006. http://hdl.handle.net/10204/1038 . en_ZA
dc.identifier.ris TY - Conference Presentation AU - Van der Walt, Christiaan M AU - Barnard, E AB - The relationship between the distribution of data, on the one hand, and classifier performance, on the other, for non-parametric classifiers has been studied. It is shown that predictable factors such as the available amount of training data (relative to the dimensionality of the feature space), the spatial variability of the effective average distance between data samples, and the type and amount of noise in the data set influence such classifiers to a significant degree. The methods developed here can be used to gain a detailed understanding of classifier design and selection. DA - 2006-11 DB - ResearchSpace DP - CSIR KW - Data classifier performance KW - Datasets KW - Non-parametric classifiers LK - https://researchspace.csir.co.za PY - 2006 T1 - Data characteristics that determine classifier performance TI - Data characteristics that determine classifier performance UR - http://hdl.handle.net/10204/1038 ER - en_ZA


Files in this item

This item appears in the following Collection(s)

Show simple item record