A Study Of Cancer-related Gene Classification Prediction Based On Cloud Model

With the continuous development of science and technology and post-genomicera, the understanding of human genes is more in-depth, but also greatly improvedmeans of detection of gene expression data and detection technology, allowsresearchers in a shorter time with less the number of experiments to obtain a largenumber of gene expression data.These data for the study of the pathogenesis ofvarious diseases, disease diagnosis, as well as the development of new drugs andgene therapy of the disease at the genetic level are of great significance.Cancer is the major diseases that affect human health. Classification andprediction of cancer-related genes in the gene level is to understand the pathogenesisof cancer, to find gene expression data of the relationship between change andcancer pathological features, in order to develop new drugs for specific genes key tothe treatment of cancer steps.However, only a small number of samples in a flood ofgenetic data can be analyzed, which resulted in serious " dimension disaster ",because a large number of cancer genetic data of unavailable, resulting in a seriousdecline of the classification performance and accuracy.In order to solve the aboveproblem, this paper will use the classification of cancer-related gene classifier basedon cloud model, application cloud model theory on cancer-related genes to classifythe literature is still rare, the paper aims at the use of cloud model in the data miningadvantages, combined with particle swarm optimization algorithm, classification andprediction of cancer-related genes.In this paper, are as follows:(1) a detailed summary of biological information and described, including the definition of biological information, resulting in the development of research areasand the main achievements of the recent research.(2) analysis and research on the hot issues of the current bioinformatics research-cancer-related gene classification and prediction problems.including theoccurrence and development of cancer and cell cycle between the gene and itsbiological significance of this application of the data set characteristics, as well asdomestic and international cancer-related gene research progress.(3) cloud model theory in detail, including the definition of the cloud, thebasic characteristics of the cloud model and cloud model three basic numericalcharacteristics.Analysis and discussion of the cloud model generator and its associatedalgorithms, and in recent decades to study the progress of the cloud model theoryare introduced.(4) particle swarm optimization algorithm with the cloud model combinedapplication of cloud model classifier Classification and prediction of cancer-relatedgenes, based on cloud classification model with other similar functions. Acomparative study and analysis of the respective the advantages and disadvantages,and propose improvements.Simultaneous analysis of different cloud classifier for avariety of applications of different algorithms in the classification effectiveness andclassification efficiency, it is verified by comparing the predictive validity of thecloud model of cancer-related gene classification based on particle swarm.