Studies in Theory and Applications of DEA for Forecasting Purpose

2015 Fiscal Year Final Research Report

• Project/Area Number: 25282090
• Research Category: Grant-in-Aid for Scientific Research (B)
• Allocation Type: Partial Multi-year Fund
• Section: 一般
• Research Field: Social systems engineering/Safety system
• Research Institution: National Graduate Institute for Policy Studies
• Principal Investigator: Tone Kaoru 政策研究大学院大学, 政策研究科, 名誉教授 (00051235)
• Co-Investigator(Kenkyū-buntansha): FUKUYAMA Hirofumi 福岡大学, 商学部, 教授 (80218958)
• Project Period (FY): 2013-04-01 – 2016-03-31
• Keywords: DEA / Forecast / Resampling / Entropy / Dynamic DEA

Outline of Final Research Achievements

(1) We developed a method for evaluating representative forecasting methods by means of DEA. Our method uses multiple criteria whereas most existing methods employ a single criterion. We applied the method to forecasting WTI crude oil price.(2) We developed a new method for evaluating confidential intervals of DEA scores. We can apply this method for past-present time series data as well as past-present-future data. (3) We employ the entropy maximization model for forecasting volatile future scenario and evaluate the past-present-future overall efficiency scores by applying Dynamic DEA models.

Free Research Field

オペレーションズリサーチ