2014 Fiscal Year Final Research Report
A unifying approach to the shape and change-point hypotheses based on doubly-accumulated statistics and its various applications
Project/Area Number |
23500362
|
Research Category |
Grant-in-Aid for Scientific Research (C)
|
Allocation Type | Multi-year Fund |
Section | 一般 |
Research Field |
Statistical science
|
Research Institution | Meisei University |
Principal Investigator |
HIROTSU Chihiro 明星大学, 公私立大学の部局等, 主幹研究員 (60016730)
|
Project Period (FY) |
2011-04-28 – 2015-03-31
|
Keywords | 1・2・3重累積和 / 1母数指数分布族 / 形状制約仮説 / 統計的品質管理 / 不等間隔時系列 / 変化点解析 / マルコフ性 / 用量反応曲線外挿 |
Outline of Final Research Achievements |
A unifying approach is developed to the shape and change-point hypotheses which have been developed in two different streams of statistics. Further the monotone, convex and sigmoid hypotheses are dealt with in a systematic way. They are corresponding to step, slope and inflection change-point models, respectively. It is also important in practice, such as monitoring of drug adverse events and the dose-response analysis, to know the change-point as well as the outline of shape. As the basic distribution a univariate exponential family is generally dealt with. Thus this is really a unifying approach to the shape and change-point hypotheses.
|
Free Research Field |
統計科学
|