DNA序列的分类

本文对 A题中给出的 DNA序列分类问题进行了讨论 .从“不同序列中碱基含量不同”入手建立了欧氏距离判别模型 ,马氏距离判别模型以及 Fisher准则判定模型 ;又从“不同序列中碱基位置不同”入手建立了利用序列相关知识的相关度分类判别算法 ,并进一步研究了带反馈的相关度分类判别算法 .对于题中所给的待分类的人工序列和自然序列 ,本文都一一作了分类 .接着 ,本文又对其它各种常见的分类算法进行了讨论 ,并着重从分类算法的稳定性上对几种方法作了比较 .

Classification of DNA Sequences

This paper proposes several methods for the classification of DNA sequences. We noticed that different sequences have different alkali radicals and therefore set up models using Euclidean distance， Mahalanobis distance and Fisher principle. We also noticed that different sequences have different permutations of alkali radicals and an algorithm using relativity analysis is proposed. Further we discussed a relativity analysis algorithm with feed-back mechanism. As to the natural and artificial data given our algorithms work well and fine results are given. At last several other common algorithms are compared， especially on their stabilities.