Mixed response surface models and Bayesian analysis of variance components for electrically conductive adhesives

作者：Berni Rossella^*; Scarano Valeria Leonarda; Bertocci Francesco; Catelani Marcantonio

来源：Applied Stochastic Models in Business and Industry, 2013, 29(4): 387-398.

DOI：10.1002/asmb.1978

• mixed-level fractional factorial design
• mixed response surface models
• variance components
• Bayesian analysis
• electrically conductive adhesives (ECAs)

This paper deals with an analysis of random effects for microelectronic data. More precisely, by considering the technical challenges related to the use of electrically conductive adhesives such as soldering material in electronics, the sources of variabilities related to different electrically conductive adhesive characteristics and working process variables are evaluated. Random effects are involved in a response surface methodology setting, and the results are compared with a Bayesian approach where variance components are estimated through a log-posterior expressed as the product of the information matrix and restricted maximum likelihood estimates of variance components.

• 出版日期2013-7