Title :
Comparing and analysing three kinds of conformal predictor
Author :
Zeng, Jian-Xin ; Gong, Hui-Li ; Liu, Ting ; Men, Yue ; Yang, Ning
Author_Institution :
Inf. Networks Dept., Hongta Tobacco (Group) Co., Yuxi, China
Abstract :
This paper studies the conformal predictor based on k-nearest neighbour, the conformal predictor based on support vector machine, the conformal predictor based on logistic regression model. We use the three conformal predictors to predict the irises standard data sets and cigarette strength index data sets, then by comparing the results we find that in irises standard data sets, the conformal predictor based on k-nearest neighbour and the conformal predictor based on support vector machine perform well. In cigarette strength index data sets the conformal predictor based on support vector machine is the best. Through the experimental analysis, we find that the result is reliable.
Keywords :
data analysis; learning (artificial intelligence); pattern classification; regression analysis; support vector machines; cigarette strength index data sets; conformal predictor; experimental analysis; irises standard data sets; k-nearest neighbour; logistic regression model; support vector machine; Accuracy; Distributed databases; Indexes; Logistics; Predictive models; Standards; Support vector machines; conformal predictor; nonconformity measure; transduction;
Conference_Titel :
Fuzzy Systems and Knowledge Discovery (FSKD), 2012 9th International Conference on
Conference_Location :
Sichuan
Print_ISBN :
978-1-4673-0025-4
DOI :
10.1109/FSKD.2012.6233751