DocumentCode
3335869
Title
A novel approach towards Naïve Bayesian classifier
Author
Singh, Bhupesh Kumar ; Agarwal, Anupam
fYear
2011
fDate
8-10 Dec. 2011
Firstpage
1
Lastpage
4
Abstract
Bayesian Theory has been an elegant basis for the design of classifiers. There are several classifiers which do not adhere with the Bayesian principle strictly but motivated from the very theory, termed as naive Bayesian methods. The present work proposes a naive Bayesian Classification approach which limits the classification process to the basic operation of the key comparison. Moreover we are able to link the two notions of the probability - frequency in ensemble and idea of reasonable expectation by the means of ordinary algebra.
Keywords
belief networks; pattern classification; probability; Bayesian principle; Naïve Bayesian classifier; ordinary algebra; probability; Accuracy; Bayesian methods; Complexity theory; Learning systems; Machine learning; Testing; Training; Naïve Bayesian Classification; Notion of Probability;
fLanguage
English
Publisher
ieee
Conference_Titel
Engineering (NUiCONE), 2011 Nirma University International Conference on
Conference_Location
Ahmedabad, Gujarat
Print_ISBN
978-1-4577-2169-4
Type
conf
DOI
10.1109/NUiConE.2011.6153232
Filename
6153232
Link To Document