چکیده مقاله
A supervised neural network using radial basis network RBN is developed The RBN uses error back propagation algorithm EBP as predictive tools for the modelling process Since NN based models are expensive techniques, Design of Experiments and statistical techniques have been employed to offset this expense A comparison between several experimental based models on predictive capability and number of training patterns is given Very often, the designer is faced with a difficult situation that sometimes information is not available In such a case, the process modeller can compromise accuracy information for the experimental cost Several 2 levels, 3 levels, 4 levels, and 5 levels OAs are used These are L8 OA, L9 OA, L27 OA, L32 OA, and L25 OA respectively Results show that each individual model has a potential for approximation if used by itself Besides an attempt to combine the models in a sequence and the resulting composed models are used and compared for approximation Results of constructing different composed models indicate that using a certain sequence leads to a better model with faster convergence and less predictive error
کلیدواژهها
نویسندگان
M.H Gadallah
Associate Professor, OperationsResearch Group, Institute of Statistical Studies & Research, Cairo University, Egypt ۱۲۶۱۳
K El-Sayed
M Sc Student, Department of Mechanical Engineering, The American University in Cairo, Egypt
K Hekman
Associate Professor, Department of Mechanical Engineering, The American University in Cairo, Egypt
شیوه ارجاع
Gadallah, M.H and El-Sayed, K and Hekman, K,1384,Radial Basis Neural Network Models: Model Development and Validation,01st Tehran International Congress on Manufacturing Engineering (TICME2005),Tehran
ارائهشده در
مجموعه مقالات اولین کنفرانس بین المللی و هفتمین کنفرانس ملی مهندسی ساخت و تولید1 فروردین 1384 · تهران