Twin Support Vector Machine Based On Novel Swarm Intelligent Optimization

Twin support vector machine is a machine learning algorithm based on support vector machine.Distinct from support vector machine,twin support vector machine finds two classification hyperplanes which are not parallel,so that each kind of sample is close to one hyperplane as far as possible and far away from another hyperplane.Twin support vector machine transforms a single large-scale quadratic programming problem into two smaller quadratic programming problems,which makes the computing speed of twin support vector machine about 4 times faster than that of support vector machine.XOR problem can be solved by twin support vector machine,and its classification performance is better than support vector machine.Although the development time of twin support vector machines is relatively short,due to its solid theoretical foundation and excellent performance,it has become a research hotspot in the field of machine learning.Although the research of twin support vector machines has made great progress in algorithm improvement and application in recent years,there are still some shortcomings in parameter selection and operation efficiency.In response to these problems,this paper studies twin support vector machine,the specific contents are as follows:In view of the twin support vector machine can't select the right parameters,the inappropriate parameters will reduce the classification ability.This paper proposed a twin support vector machine based on artificial fish swarm algorithm(AFSA-TWSVM),which combines artificial fish swarm algorithm and twin support vector machine to solve the problem of difficult parameter selection.First,the parameters of the support vector machine are taken as the position information of the artificial fish,and the classification accuracy is taken as the objective function.Then,the position and the optimal solution are updated by the artificial fish's preying,swarming,following,and random behavior.At the end of the iterations,the optimal parameters and the optimal classification accuracy are obtained.Experiments on datasets show that the algorithm can avoid the blindness of parameter selection,automatically determine parameters in the process of model training,and improve the classification performance of twin support vector machines.In order to accelerate the speed of parameter selection,this paper introduces a new group intelligent optimization algorithm to optimize the twin support vector machine,and proposes a twin support vector machine based on butterfly optimization algorithm(BOA-TWSVM).The parameters of the twin support vector machine are searched by the constant iteration of butterfly's automatic search behavior.In the process of searching,the switching probability is used to select the global searching behavior and the local random walking behavior,so as to increase the searching speed and ability.Finally,the optimal model of twin support vector machines is established according to the optimal solution.The experimental results on UCI dataset show that BOA-TWSVM not only performs well in classification accuracy,but also accelerates the speed of parameter selection.