CAREER: Statistical Search Techniques for Human Genome and Computer Chess

职业：人类基因组和计算机国际象棋的统计搜索技术

• 批准号: 9702071
• 负责人: Thomas Anantharaman
• 金额: $ 22.2万
• 依托单位: New York University
• 依托单位国家: 美国
• 项目类别: Continuing Grant
• 财政年份: 1997
• 资助国家: 美国
• 起止时间: 1997-06-01 至 2001-07-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=9702071&HistoricalAwards=false
• 关键词: CAREER; Statistical; Search; Techniques; Human

The goal of this research project is to explore the combination of statistical modeling with heuristic search techniques to solve hard computational problems in two areas : construction of a physical map of the human DNA and automatic learning in computer chess. The approach consists of formulating statistical models that are trained on data to automatically adjust broad parameters that define the knowledge encoded in the program. This is combined with traditional AI heuristic search techniques when exhaustive search of the parameter or problem space is not possible. The results of this research will allow programs to exhibit a rudimentary form of learning in areas like chess games, as well as allow a physical map of the entire human DNA to be constructed in record time of about six months which will allow geneticists to find the cause of any genetic disorder and in turn increases the chance of finding cures for these genetic disorders.

该研究项目的目标是探索统计建模与启发式搜索技术的结合，以解决两个领域的计算难题：构建人类DNA的物理图谱和计算机象棋中的自动学习。该方法包括制定统计模型，这些模型在数据上进行训练，以自动调整定义知识的广泛参数 编码在程序中。当参数或问题空间的穷举搜索不可能时，这与传统的AI启发式搜索技术相结合。 这项研究的结果将允许程序在国际象棋游戏等领域表现出基本的学习形式，并允许在大约六个月的创纪录时间内构建整个人类DNA的物理图谱，这将使遗传学家能够找到任何遗传疾病的原因，从而增加找到治愈这些遗传疾病的机会。