Statistical modeling of spatio-temporal data and quantitative grasp of phenomena

2014 Fiscal Year Final Research Report

• Project/Area Number: 23300106
• Research Category: Grant-in-Aid for Scientific Research (B)
• Allocation Type: Single-year Grants
• Section: 一般
• Research Field: Statistical science
• Research Institution: Kyushu University
• Principal Investigator: NISHII Ryuei 九州大学, マス・フォア・インダストリ研究所, 教授 (40127684)
• Co-Investigator(Kenkyū-buntansha): NINOMIYA Yoshiyuki 九州大学, マス・フォア・インダストリ研究所, 准教授 (50343330) ; MASUDA Hiroki 九州大学, マス・フォア・インダストリ研究所, 准教授 (10380669) ; MAESONO Yoshihiko 九州大学, 大学院数理学研究院, 教授 (30173701) ; TANAKA Shojiro 島根大学, 大学院総合理工学研究科, 教授 (00197427) ; SHIMIZU Kunio 慶應義塾大学, 名誉教授 (60110946) ; QIN Pan 九州大学, マス・フォア・インダストリ研究所, 学術研究員 (40532718) ; KONISHI Sadanori 中央大学, 理工学部, 教授 (40090550) ; SAKATA Toshio 九州大学, 大学院芸術工学研究院, 教授 (20117352)
• Project Period (FY): 2011-04-01 – 2015-03-31
• Keywords: 時空間モデリング / 時系列解析 / 環境統計 / ノンパラメトリック解析 / 森林減少 / 宇宙天気 / 確率微分方程式 / モデル選択

Outline of Final Research Achievements

The aim of our research was to promote statistical modeling for spatio-temporal data acquired by actual phenomena, and grasp its characteristics through the modeling. Also we studied statistical properties of the models. The concrete results are as follows.
1) We considered deforestation models where forest-covering-ratio is regressed by human population density and relief energy with spatial and temporal dependency. 2) Auto-regressive models with exogenous variables are extended through generalized linear models, and applied to actual data. Then, magnetic storm time index is accurately predicted by using physical measurements of solar activities. 3) Frequencies of weak earthquakes were shown to be affected by the solar wind. 4) Asymptotic theory for the parameter estimation of time series models described by stochastic differential equations was developed.

Free Research Field

統計学