标题: 逼近演算法之设计与分析
On the Design and Analysis of Approximation Algorithms
作者: 蔡锡钧
TSAI SHI-CHUN
国立交通大学资讯工程学系(所)
公开日期: 2014
摘要: 演算法自计算机科学发展以来一直是核心研究中的一部分,而其中最佳化问题算法的 演进也都是重要的议题。本计昼的三个主题在:

一、为特定的算法问题设计有效率的逼近算法或进行average case complexity的分 析:

在P关NP几近成立的前提下,许多属于NP-Hard但重要而常用来建模的算法 问题要有效率的完美求解已是希望渺茫。进而解决此类问题的方向便是进行逼近 算法的研究或者分析在average case下,问题是否会变得容易许多。面对有多项 式时间算法但仍希望增进时间效率的情况,也可以尝试放宽求解的质量进而设计 高效算法。

二、分析问题的不可逼近性:

面对逼近算法也难以设计的情况下,必须从另一方面探讨问题可逼近的极限,是 否连逼近本身也是真正困难的。

三、问题的计算复杂度归类:

许多问题虽然可能已知是NP-Hard,但详细应被归类在哪还未有结果,在对于算法 问题的算法设计与结构上的仔细分析后会对问题的本质有更多的了解,有助于我 们将尚未归类的难题作分类。进一步了解问题的困难度。
The design and analysis of algorithms has been the central topic of computer science, since the very beginning of this field. Although there is no proof of P^NP, we believe there are many problems that cannot be solved in polynomial time. Thus the idea of approximation algorithms comes to play. In this project we will propose to investigate the following three issues:

1. Design efficient approximation algorithms for specific optimization problems and give average case analyses, etc. instead of worst case analysis.

2. Investigate the in-approximality of optimization problems. Improve the ratio of approximation of specific problems.

3. Furthermore, since we know some problems are in-approximable within some factor unless P=NP. We can study the more accurate higher order class, such as Sigma_2 or Pi_2, that the in-approximable problem should belong to.
官方说明文件#: NSC101-2221-E009-025-MY3
URI: http://hdl.handle.net/11536/102425
https://www.grb.gov.tw/search/planDetail?id=8112869&docId=430289
显示于类别:Research Plans