论文部分内容阅读
流感病毒分为三类:甲型(A型),乙型(B型),丙型(C型).在这三种类型中甲型(A型)流感病毒是最致命的流感病毒,对人类引起了严重疾病.本文对甲型流感病毒DNA序列建立了一种新的时间序列模型,即CGR(Chaos Game Representation)弧度序列.利用CGR坐标将甲流病毒DNA序列转换成CGR弧度序列,且引入长记忆ARFIMA模型去拟合此类序列,发现随机找来的10条H1N1序列,10条H3N2序列都具有长相关性且拟合很好,并且还发现这两种序列可以尝试用不同的ARFIMA模型去识别,其中H1N1可用ARFIMA(0,d,5)模型去识别,H3N2可用ARFIMA(1,d,1)模型去识别.
Influenza viruses fall into three categories: Type A (Type A), Type B (Type B), Type C (Type C). Of these three types, Influenza A (A) is the deadliest influenza virus, Human caused a serious disease.This paper established a new time series model of influenza A (DNA) sequence of influenza virus, which is CGR (Chaos Game Representation) radian sequence.The CGR coordinates were used to convert the DNA sequence of influenza A virus into CGR radians and The long memory ARFIMA model was introduced to fit these sequences. The 10 H1N1 and 10 H3N2 sequences were found to be highly correlated and fit well, and the two sequences were found to be able to try different ARFIMA H1N1 can be identified by ARFIMA (0, d, 5) model and H3N2 can be identified by ARFIMA (1, d, 1) model.