Cooperative USA-Latvia Research

该奖项资助了美国和拉脱维亚研究人员之间的合作研究努力。具体地说，该项目资助一名拉脱维亚研究人员访问马里兰大学与国际和平研究所进行合作研究的旅费。获奖期为两年。这项研究考察了机器学习领域的问题。该项目研究假设空间的相对大小，学习算法必须在该空间中搜索才能找到正确的解决方案。本研究采用了几种新的样例学习观点进行研究。该项目还推广了正则表达式，这是术语重写系统的扩展，已经为其实现了实用的学习算法。该项目开发了将这些实现扩展到更广泛的功能类的技术。

This award funds a cooperative research effort between US and Latvian researchers. Mores specifically, this project supports the travel expenses of a Latvian researcher to visit the University of Maryland for the collaborative research with the PI. The award period is for two years. The research examines issues in the area of machine learning. The project studies relative sizes of the hypothesis space that a learning algorithm must search through in order to find a correct solution. Several new views of learning by examples are used for investigation. The project also generalizes regular expressions, which are extensions of the term rewriting systems for which practical learning algorithms have been implemented. The project develops techniques for extending these implementations to a broader class of functions.