BIMSA >
生物信息讨论班
Energy entropy vector: an efficient method for coding and classifying unpaired gene sequences
Energy entropy vector: an efficient method for coding and classifying unpaired gene sequences
组织者
演讲者
时间
2024年01月06日 21:30 至 22:00
地点
Online
摘要
In this study, a novel gene encoding method, energy entropy vector, is proposed. The method does not require comparison or interception, and is able to map gene sequences of arbitrary length into 18-dimensional vectors, overcoming the limitations of traditional methods in feature extraction. We performed experimental validation on five microbial datasets and compared it with the natural vector method and the covariance natural vector method. The experimental results show that the proposed method performs well in convex packet classification, kingdom, and family classification tasks, especially in the family classification task, where the accuracy is improved by 15% to 30%.