Shuffled Frog Leaping Algorithm Based Neural Network and Its Application in Big Data Set

Abstract

This paper studies the random optimization algorithm to reorganize the leapfrog algorithm. The big data challenge requires an effective optimization algorithm to explore potential data structures using deep neural networks. At first, we introduce the neural network classifier and compare it with the support vector machine. Neural networks are suitable for large data sets and have the complex ability to extract high-level abstract data. And then we have to introduce a large dataset covering cancer data and voice data. Both datasets have large numbers of samples with complex low-level variance. At last we have to use the reorganized leapfrog algorithm to optimize neural network parameters. The random leapfrog algorithm is efficient and robust to a local minimum. The experimental results show that the algorithm has extensive application prospects and is suitable for the classification of big dataset. The neural network parameters can effectively optimized by the improved shuffled frog leaping algorithm.