项目名称: 约束条件下的模型选择及相关问题的研究
项目编号: No.10871037
项目类型: 面上项目
立项/批准年度: 2009
项目学科: 金属学与金属工艺
项目作者: 张宝学
作者单位: 东北师范大学
项目金额: 29万元
中文摘要: 本项目主要研究约束条件下的模型选择问题,该问题在药物临床试验、基因表达数据分析、教育心理、可靠性等领域的理论研究和实际应用中具有十分重要的意义,是统计学研究的前沿问题之一。虽然目前该领域取得一些成果,但由于在研究方法的限制,该领域的研究一直停滞不前,如何结合实际应用背景,采取新的研究方法使得该方向的研究取得更大的突破,是本项目追求的目标,这无论在理论上还是在实际应用中都具有十分重要的意义。本项目拟研究的内容,是模型选择理论中的核心问题之一。具体研究内容是为了克服已有准则的缺陷,给出新的信息准则,在此基础上,给出分布未知时,不等式约束条件下的信息准则,并讨论准则的性质。最后,应用到具体的实际问题中去。本项目的研究可以通过提出新方法、获得新结果来丰富模型选择理论,填补模型选择理论的空白。同时,又为实际应用提供理论依据和指导。
中文关键词: 约束条件;模型选择;信息准则
英文摘要: The projects mainly research the model selection problem under restrictions, which in clinical trials, gene expression data analysis, educational psychology, reliability and other areas of theoretical research and practical application has the very vital significance, and is statistically one of the frontier topics. Although it achieved some results, but because of the research method of limits, research in this field has been remain stagnant, how to combine the practical application background, adopt new research method makes the study direction to achieve greater breakthroughs, is the pursuit of the objectives of the project, whether in theory or in actual application with very important. This project's content is one of the core issues on the model selection theory. The concrete research projects will propose new information criterion by overcoming the existing standard defects, especially, when the distribution are unknown, information criterion are given under inequality constraints and discuss its properties. Finally, the application to specific practical problems. This research can come up with new methods, new results to enrich the model selection theory, fill the blank model selection theory. At the same time, but also for practical applications to provide a theoretical basis and guidance.
英文关键词: Constraints;Model Selection;Information Criterion