CISE Research Instrumentation: Distributed Data Mining in Large Databases

点击翻译按钮获取中文摘要

9421531 Salzberg This award is to purchase equipment to be used for three distinct research projects. The first project focuses on computational biology, especially on distinguishing exons and introns in raw DNA sequence data using statistical analysis and machine learning algorithms. This project will also study algorithms for searching protein sequence databases to find fundamental building blocks called motifs. The second project is a continuation and extension of previous work on analysis of very large databases of astronomical images from the Hubble Space Telescope and ground-based telescopes. The development of automated classifiers to these databases is crucial for the success of astronomical sky surveys, in which data on hundreds of millions of celestial objects needs to be collected and classified in real time. In the third project, the research group will develop distributed versions of their corpus-based language learning algorithms, which will allow them to run many more experiments and to make much more rapid progress in this research area. The algorithms under development, which are computationally intensive, automatically acquire linguistic information from large text corpora, instead of having people manually input this information based upon their linguistic intuitions.