Effective hybrid feature selection method based on improved binary krill swarm algorithm and information gain algorithm

Abstract

The invention provides an effective hybrid feature selection method based on an improved binary krill swarm algorithm and an information gain algorithm. The algorithm comprises the following steps: step 1, randomly initializing N individuals in a population by adopting a feature sorting strategy based on an information gain algorithm; 2, calculating a fitness value of each individual in the population by adopting a set fitness function, and taking a solution expressed by the individual with the maximum fitness value in the population as a global optimal solution in the population; 3, updatingthe population by using an improved binary krill swarm algorithm, updating the fitness value of each individual in the population, and updating a globally optimal solution in the population; and 4, taking the step 3 as one iteration, and repeating the step 3 until the current number of iterations reaches the set number of iterations. Through test verification of 10-fold crossing on nine public biomedical data sets, the number of gene expression levels can be effectively reduced, and compared with other feature selection methods, high classification accuracy is obtained.

Description

technical field [0001] The invention relates to the technical field of bioinformatics, in particular to an effective mixed feature selection method based on an improved binary krill swarm algorithm and an information gain algorithm. Background technique [0002] With the development of DNA microarray technology in biomedicine, large-scale high-dimensional small-sample microarray data have been accumulated, such as the lung cancer microarray data set, including 181 samples, each containing 12533 ​​features. The high-dimensional and small-sample characteristics of the data have brought great challenges to gene analysis and disease diagnosis. High-dimensional small-sample data also brings great challenges to existing mining and learning algorithms. With the sharp increase of data dimensions, a large amount of redundant information and irrelevant information will usually be generated, which may greatly reduce the performance of machine learning algorithms, increase computationa...

AI Technical Summary

Problems solved by technology

[0012] For the existing feature selection algorithm, it is impossible to cover the two aspects of "selecting the most valuable feature subset composed of relevant features from the original input data" and "improving the classification accuracy as much as possible". To solve the target problem, the present invention provides an effective mixed feature selection method based on the improved binary krill swarm algorithm and information gain algorithm, which can further improve the classification accuracy of features while selecting the optimal feature subset

Images