Research On Support Vector Machine Based On The Expedited Artificial Fish-swarm

Now, with the arrival of big data, it has been increasingly widespread and popular to support vector machine applications. But the support vector machine itself is a complicated calculation process. And the calculation time complexity and space complexity is relatively high. In addition, there are other issues such as not easily parallelized. At the same time, the bionics algorithms such as artificial fish-swarm have also been widely used, mainly to solve the optimal solution. To solve the existing problems in support vector machine algorithm, this paper aims to merge the artificial fish-swarm algorithm into support vector machine algorithm.Firstly, we try to use artificial fish-swarm algorithm to solve the quadratic programming problems in support vector machine algorithm. With the artificial fish-swarm algorithm, we carry on the process of minimum optimization. Experimental results show that the optimized support vector machine of artificial fish swarm intelligence algorithm (AFA-SVM) has the advantages of both. It has fast convergence and parallel characteristic. Due to the nature of artificial fish swarm intelligence algorithm, when a sample set is changed, the original training set remains the reference value in order to achieve an incremental classifier.Secondly, there are many experts and scholars researching in hybrid applications of bionics intelligent algorithm. The particle swarm and fish swarm algorithm we use are derived from studies on lifestyle characteristics of birds or fish. This article makes a bold fusion of two algorithms. We make the acceleration of particle swarm apply into clusters and rear-end behavior of artificial fish swarm intelligence algorithm to achieve the expedited artificial fish swarm intelligence algorithm (EAFA)Finally, based on the improvement of support vector machine based on the artificial fish swarm, we hope to further accelerate the convergence rate. So this paper will attempt to integrate the expedited artificial fish swarm intelligence algorithm into support vector machine algorithm to achieve the EAFA-SVM algorithm. Experimental results show that EAFA-SVM algorithm compared with AFA-S VM algorithm has been significantly improved in the convergence rate.