A Combinational Approach for Experts of Committee Machine in Wire Bonding Problem Solving

Abstract:

There are three phases to solve multiple response optimization (MRO) problems. They are experiment design, modelling and optimization. Different committee machines (CMs) are applied for modeling of MRO problem. Two kinds of CM introduced by current authors in recent works. Initial approach includs utilization of all experts in CM and finally sequentilal combinational model (SCM) is created. Current study proposes a combinational approach to filter and select the best experts to create CM which yields optimum combinational model (OCM). Also , there was another modeling with name point approach model (PAM). This study investigated two different features of different SCMs, PAM and OCM in a famous MRO problem with name wire-bonding. This problem was defined in 1997 and then, several researchers solved it in next years. The results show PAM has minimum RMSE but its global desirability is low. Also overall RMSE has a slightly fall from SCMs to OCM , whereas, there is gently rise in global desirability from SCMs to OCM and consequently proposed approach or OCM provides superior results.

• Periodical: Applied Mechanics and Materials (Volumes 465-466)
• Pages: 751-755
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.465-466.751
• Online since: December 2013
• Authors: Seyed Jafar Golestaneh, Ismail Napsiah, Mohd Khairol Anuar Bin Mohd Ariffin, Say Hong Tang, Hassan Moslemi Naeini, Ali Asghar Maghsoudi, Zahra Firoozi
• Keywords: Committee Machine, Multiple Responses Optimization, Wire-Bonding Process

References

[1] S. J. Golestaneh, N. Ismail, S. H. Tang, M. K. A. M. Ariffin, H. M. Naeini, A. A. Maghsoudi, A committee machine approach to multiple response optimization, International Journal of the Physical Sciences, 6 (2011) 7935 - 7949.

DOI: 10.4028/www.scientific.net/amm.564.608

[2] S. J. Golestaneh, N. Ismail, S. H. Tang, M. K. A. M. Ariffin, H. M. Naeini, Improvement of Committee Machine Performance to Solve Multiple Response Optimization Problems, International Journal of Engineering & Technology, 13 (2013) 52-63.

DOI: 10.1155/2013/628313

[3] H. -H. Chang, Y. -K. Chen, Neuro-genetic approach to optimize parameter design of dynamic multiresponse experiments, Appl Soft Comput, 11 (2011) 436–442.

DOI: 10.1016/j.asoc.2009.12.002

[4] N. Ismail, S. J. Golestaneh, S. H. Tang, M. K. A. M. Ariffin, H. Moslemi Naeini, D. Rouhani. Modified committee neural networks for prediction of machine failure times. The 3rd national intelligent systems and information technology symposium (ISITS 2010). (2010).

[5] A. Kadkhodaie-Ilkhchi, M. R. Rezaee, H. Rahimpour-Bonab, A committee neural network for prediction of normalized oil content from well log data: An example from South Pars Gas Field, Persian Gulf, Journal of Petroleum Science and Engineering 65 (2009).

DOI: 10.1016/j.petrol.2008.12.012

[6] T. Kamo, C. Dagli, Hybrid approach to the Japanese candlestick method for financial forecasting, Expert Syst Appl, 36 (2009) 5023-5030.

DOI: 10.1016/j.eswa.2008.06.050

[7] Matlab User's Guide, Version 2012b, Neural Network Toolbox Math Works, USA, (2012).

[8] H. B. Celikoglu, Application of radial basis function and generalized regression neural networks in non-linear utility function specification for travel mode choice modelling, Math Comput Modell, 44 (2006) 640–658.

DOI: 10.1016/j.mcm.2006.02.002

[9] E. Ardil, P. S. Sandhu, A soft computing approach for modeling of severity of faults in software systems, International Journal of Physical Sciences, 5 (2010) 074-085.

[10] R. Noorossana, S. D. Tajbakhsh, A. Saghaei, An artificial neural network approach to multiple-response optimization, Int J Adv Manuf Technol, (2008) doi: 10. 1007/s00170-00008-01423-00177.

DOI: 10.1007/s00170-008-1423-7

[11] E. Del Castillo, D. C. Montgomery, D. R. Mccarviille, Modified desirability function for multiple response optimization, Journal of quality technology, 28 (1996) 337-345.

DOI: 10.1080/00224065.1996.11979684

[12] Z. He, P. F. Zhu. A hybrid genetic algorithm for multiresponse parameter optimization within desirability function framewor. 16th international conference of Industrial engineering and engineering management, IEEE. (2009).

DOI: 10.1109/icieem.2009.5344518