SGER GOALI: Improving conjoint-based consumer preference models for use in engineering design optimization

这一SGER目标项目旨在通过开发准确的消费者对新产品设计需求的分析模型来改进设计优化实践。探索性活动将探讨使用现有的基于市场研究的联合分析方法来开发可整合到设计优化中的消费者偏好模型的可行性。风险在于，除了最简单的建模外，这种模型是否可以扩展到工程领域。复杂的产品设计需要创建可验证的估计算法。只有在建模的属性数量增加的同时减少估计时间，这才是可行的。能否开发出一种稳健的方法，从而满足SGR所需的高风险组件，这是一个高度值得怀疑的问题。更广泛的影响来自于在汽车和计算机等消费产品的设计优化方面实现改进的高潜力。如果这种方法被证明是可行的，它开辟了新的研究途径，可以满足相对于客户需求对产品的改进，同时也可以推进预测产品实现的验证方法。这个SGER是一个目标项目，通用汽车作为直接合作伙伴，通过智力参与和资金投入。通用汽车将带来消费者偏好建模的贝叶斯估计方法方面的专业知识，以及用于验证这一探索性研究目的的数据

This SGER GOALI project aims to improve the practice of design optimization by developing accurate analytical models of consumer demand for new product designs. The exploratory activity will address the feasibility of using an existing market research based method of conjoint analysis to develop consumer preference models that can be integrated into design optimization. The risk is whether such models can be extended into the engineering domain for anything but the most simple modeling. Complex product design requires the creation of estimation algorithms that can be validated. This will only be feasible if the estimation time can be reduced while the number of attributes being modeled is increased. It is highly questionable if a robust method can be developed, thus satisfying the high-risk component necessary for an SGER.The broader impacts come from the high potential of attaining improvements in design optimization of consumer products such as automobiles and computers. If this methodology proves feasible, it opens up new research avenues where improvements to the product relative to customer needs can be met, while the validation methodology for predictive product realization can also be advanced. This SGER is a GOALI project with GM as a direct partner, through both intellectual participation and funding. GM will bring expertise in Bayesian estimation methods of consumer preference modeling as well as data for validation purposes for this exploratory research