2010年研究生数学建模竞赛A题综述

The Summarization of the Problem A

第七届全国研究生数学建模竞赛A题是生物信息学中的一个急需解决的问题.虽然有关问题的研究已经经历了十多年,但由于问题的复杂性,人们的认识还很局限,基本的结论大多还以定性的为主,定量的探讨正方兴未艾.对参赛队员来讲解决该问题是一个极大的挑战.研究生们在讨论该问题时,大多直接进行分类.然而对于一个小样本的学习问题,显然这样做是行不通的.所以问题的关键是从数学和生物学角度减少用于分类的特征数目.同时,对于获取的基因标签,需要从临床上或生物学角度找到验证.该问题的求解过程引导研究生们从数学建模走向解决实际问题.

The problem A of the seventh national mathematical modeling contest for grad- uate students came from a real research projection of bioinformatics. Since it is a complex problem, which has been studied for more than ten years, we have little knowledge about the solution of the problem. The mostly exist result about the problem is qualitative. The search for gene marker from quantitative view point is studying. It is a big challenge for the participate students to solve such real problem. When trying to solve the problem, most of graduate students directly used the classic classify method. However, it is a learning problem from small samples. The key problem is how to reduce the number of the features for classifying from the combination of mathematics and biology. In the meantime, for the found gene markers, we need to identify them from the view point of the clinic or biology. The procedure of solving the problem will lead the graduate students learning the ability of solving the real problem from the mathematics modeling contest.