Econometric investigation of random coefficient logit models of demand and its extension to multi-period demand estimation

估计差异化商品的需求对于评估品牌资产和定价至关重要，这两个领域是营销中最重要的两个领域。通常情况下，没有一组家庭的购买历史，而只有研究中的产品的市场份额和特征。通过这项基金资助的研究，我们提出了基于频数和贝叶斯理论的需求和供给的随机系数离散选择模型的估计方法，并为将该模型扩展到多阶段的选择问题做了准备。

Estimating demand for differentiated goods is essential in evaluating brand equity and pricing, two of the most important areas in marketing. It is often the case that purchasing history of a group of households is not available, but instead only the market share and characteristics of products under study are available. Through the research funded by this grant, we proposed method to estimate models of great importance, random-coefficients discrete choice models of demand and supply via frequentist and Bayesian perspectives and made preparations to expand the model to choice problems over multiple periods.