Studies on Statistical Inference and Experimental Designs Based on Algebraic Structures

2016 Fiscal Year Final Research Report

• Project/Area Number: 15H06531
• Research Category: Grant-in-Aid for Research Activity Start-up
• Allocation Type: Single-year Grants
• Research Field: Statistical science
• Research Institution: Tokyo Metropolitan University
• Principal Investigator: Ogawa Mitsunori 首都大学東京, 社会科学研究科, 助教 (50758290)
• Project Period (FY): 2015-08-28 – 2017-03-31
• Keywords: 統計計算 / 代数統計

Outline of Final Research Achievements

In statistical inference, the computation of complicated integrals or large-size summation sometimes makes the problem intractable. To overcome such problems, the numerical computation method, called the holonomic gradient method that utilize the differential equations was proposed and applied to many statistical problems. In this project, we gave the concrete calculation procedure for the holonomic gradient method using GKZ-hypergeometric system associated with Veronese configurations. We also studied the problem of experimental design. Some results on designs of variable resolutions were extended to the case of multi-level designs.

Free Research Field

計算代数統計学