Algorithms for Metric Space Embeddings

CCR-9820879 Farach-Colton操作度量空间是许多领域的基本任务。 聚类、数据可视化和最近邻搜索是几乎所有数据密集型领域中出现的一些标准原语。 解决这些问题的最有用的统一方法之一涉及嵌入的概念。 在嵌入中，一个输入度量被映射到另一个度量，对于这个度量已经开发了有用的算法。这个项目考虑度量空间嵌入中的基本算法问题。 从图像数据库索引、蛋白质数据库索引和数据库建设等方面总结出了一些问题。 然而，研究的大多数问题是基本的算法问题。

CCR-9820879Farach-ColtonManipulating metric spaces is a basic task in many areas. Clustering, data visualization, and nearest neighbor searches are some of the standard primitives which come up in almost any data-intensive domain. One of the most useful unifying approaches to addessing these problems involves the notion of an embedding. In an embedding, an input metric is mapped into another metric for which useful algorithms have been developed.This project considers basic algorithmic problems in metric space embeddings. Some of the problems were abstracted from work in image database indexing, protein database indexing and phylogency construction. However, most problems studied are basic algorithmic questions.