Feature Selection and Data Classification of Binary Rat Swarm Optimization Algorithm
In view of the difficulty of improving classification accuracy and reducing the number of feature selection in feature selection technology,with the increase of data dimension,this paper improves the new bionic optimization algorithm mouse swarm optimization algorithm,introduces the conversion function into the algorithm,uses the K-nearest neighbor method as the classifier,and proposes a binary mouse swarm optimization algorithm for feature selection and data classification,effectively reduces the di-mension of features and the error rate of data classification.It is tested on 10 data sets of UCI and compared with genetic algorithm,particle swarm optimization algorithm,bottle sea squirt swarm optimization algorithm and sine cosine algorithm.The simulation re-sults show that the algorithm can improve the accuracy of data classification and effectively reduce the feature dimension.The algo-rithm has good convergence and robustness.
rat swarm optimizationfeature selectiondata classificationKNN
国家科技期刊平台
NSTL
万方数据
