Title :
A learning method for small data sets with multimodality variables
Author :
Der-Chiang Li ; Yu-Ching Chang ; Mei-Lan Su ; Liang-Sian Lin
Author_Institution :
Dept. of Ind. & Inf. Manage., Nat. Cheng Kung Univ., Tainan, Taiwan
Abstract :
This paper proposes a new approach to generate virtual sample for learning small data set considering a distribution with multimodality. This modality test strategy is mainly aimed at assessing data clustering by using a hypothesis testing based on a uni-modality function. One data set is provided to check the effectiveness of the proposed approach.
Keywords :
learning (artificial intelligence); pattern clustering; data clustering; hypothesis testing; learning method; modality test strategy; multimodality variables; unimodality function; Accuracy; Density functional theory; Educational institutions; Sociology; Statistics; Testing; Weibull distribution; Multimodality; Small data sets; Virtual sample generation;
Conference_Titel :
Grey Systems and Intelligent Services, 2013 IEEE International Conference on
Conference_Location :
Macao
Print_ISBN :
978-1-4673-5247-5
DOI :
10.1109/GSIS.2013.6714832
