Bayesian Inference, Modeling and Forecasting Techniques in Econometrics

计量经济学中的贝叶斯推理、建模和预测技术

• 批准号: 9012328
• 负责人: Arnold Zellner
• 金额: $ 1.4万
• 依托单位: University of Chicago
• 依托单位国家: 美国
• 项目类别: Continuing Grant
• 财政年份: 1990
• 资助国家: 美国
• 起止时间: 1990-11-15 至 1992-01-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=9012328&HistoricalAwards=false
• 关键词: Bayesian; Inference; Modeling; Forecasting; Techniques

This project continues research on producing improved and new Bayesian econometric methods and applying them in analyzing and predicting economic data. Topics to be pursued include: (1) work on Bayesian learning models and procedures for generating prior densities and models for observations; (2) Bayesian model and forecast combining techniques with applications; and (3) macroeconomic theory and Bayesian macroeconometric forecasting equations. This research should improve existing methods of econometric inference and estimation. The theoretical derivations of information processing rules could lead to a powerful new framework for analyzing statistical information. The Bayesian diagnostic checking procedures developed in this project will prove useful in detecting outliers and departures from specifying assumptions in econometric studies. A predictive test of when to combine and when not to combine forecasts developed under the previous NSF grant will be generalized and applied to economic problems.

该项目继续研究生产改进和新的贝叶斯计量经济学方法，并将其应用于分析和预测经济数据。 要追求的主题包括：（1）在贝叶斯学习模型和过程中的工作，以产生先前的密度和观测模型； （2）贝叶斯模型和预测将技术与应用结合； （3）宏观经济理论和贝叶斯宏观经济学方程。 这项研究应改善现有的计量经济学推断和估计方法。 信息处理规则的理论推导可能会导致一个有力的新框架来分析统计信息。 该项目中开发的贝叶斯诊断检查程序将被证明可用于检测到计量经济学研究中指定假设的异常值和偏离。 对何时合并以及何时不合并前NSF赠款制定的预测的预测测试将被概括并应用于经济问题。