Title :
Balanced Resampling for Neural Model Selection
Author :
Hung, Wen-Liang ; Chuang, Shun-Chin
Author_Institution :
Grad. Inst. of Comput. Sci., Nat. Hsinchu Univ. of Educ., Hsinchu, Taiwan
Abstract :
In this paper we apply the balanced resampling, which is an efficient bootstrap technique for neural model selection. Our goal is to reduce computer time, so that in the bootstrap procedure, resampling is not done uniformly, this distribution is modified to obtain variance reduction. Efficiency property of this alternative distribution is shown, together with numerical data.
Keywords :
bootstrapping; neural nets; sampling methods; balanced resampling; bootstrap technique; neural model selection; Backpropagation; Computer networks; Computer science; Computer science education; Distributed computing; Logistics; Multi-layer neural network; Multilayer perceptrons; Neural networks; Pattern recognition; Asympotic relative efficiency; Balanced resampling; Bootstrap; Neural model selection;
Conference_Titel :
Natural Computation, 2009. ICNC '09. Fifth International Conference on
Conference_Location :
Tianjin
Print_ISBN :
978-0-7695-3736-8
DOI :
10.1109/ICNC.2009.586
