Gene classification method and related equipment

Abstract

The invention discloses a gene classification method and related equipment. The method comprises the steps of carrying out standardization processing on input gene data serving as training samples, wherein the gene data comprises a plurality of attributes; by difference values between intervals between the training samples and heterogeneous neighbor samples and intervals between the training samples and similar neighbor samples, and weight vectors of the corresponding attributes, building a logical regression optimization function which represents expected interval difference of all the training samples, building a minimized optimization model by taking the weight vectors of the corresponding attributes as norm constraint terms, and performing calculation through iterative operation to obtain the weight vectors of the corresponding attributes; and sorting attribute features according to values of the obtained weight vectors, carrying out classification training on the gene data servingas the training samples according to the sorted attribute features, and obtaining an optimal feature set as a classification basis, thereby classifying the to-be-classified gene data. According to the gene classification method and the related equipment provided by the invention, the relatively high classification precision can be obtained.

Description

technical field [0001] The invention relates to the technical field of data processing, in particular to a gene classification method and system. The invention also relates to a gene classification device and a computer-readable storage medium. Background technique [0002] In recent years, the research on cancer diseases has attracted the attention of more and more researchers in the fields of medicine and biology. Among them, DNA microarray technology can study cancer diseases from the perspective of genes, and is one of the currently used research methods. DNA microarray (DNA microarray), also known as gene chip, is distributed on a special substrate with DNA microarray coating. After one test, the expression data of thousands of genes can be determined, providing a data basis for disease research. [0003] After obtaining a large amount of gene expression data, it is necessary to select genes carrying important disease information, that is, to classify genes according ...

AI Technical Summary

Problems solved by technology

However, since the studied cancer gene expression data samples are few, and each sample contains tens of thousands of gene dimensions, this method requires a large amount of computing resources; moreover, only some of these genes are pairs of genes The classification problem plays a decisive role

These problems have made it difficult to adopt traditional machine learning methods, and cannot obtain satisfactory classification results.

Images