Parallelizing Gene Expression Programming Algorithm in Enabling Large-Scale Classification

Abstract

As one of the most effective function mining algorithms, Gene Expression Programming (GEP) algorithm has been widely used in classification, pattern recognition, prediction, and other research fields. Based on the self-evolution, GEP is able to mine an optimal function for dealing with further complicated tasks. However, in big data researches, GEP encounters low efficiency issue due to its long time mining processes. To improve the efficiency of GEP in big data researches especially for processing large-scale classification tasks, this paper presents a parallelized GEP algorithm using MapReduce computing model. The experimental results show that the presented algorithm is scalable and efficient for processing large-scale classification tasks.

Implementing Gene Expression Programming in the Parallel Environment for Big Datasets’ Classification

“…As it has been observed by Xu et al 15 in big data research, GEP encounters low e±ciency issue due to its long time mining processes. To improve the e±ciency of GEP, their paper proposes a parallelized GEP algorithm using MapReduce computing model.…”

“…In Ref. 15, 60 instances from the original Wine dataset have been used as the testing set, the rest was multiplied to 1,024 MB and was used as the training data. A parallel GEP-based algorithms classi¯ed data in about 5,000 s. The experiment was run on a cluster with¯ve nodes.…”

MapReduce‐based parallel GEP algorithm for efficient function mining in big data applications

“…These previous works motivate two types of MapReduce‐based distributed GEP algorithms presented in this paper. The first distributed GEP algorithm based on our previous work specially focuses on processing the large‐scale classification. However, the first algorithm cannot directly output the mined equation.…”

“…This point significantly leads to the accuracy loss of the mined equation in each mapper. In order to complement the accuracy for the further parallelized classification, the ensemble techniques with bootstrapping and majority voting are adopted …”

“…In order to complement the accuracy for the further parallelized classification, the ensemble techniques with bootstrapping and majority voting are adopted. 16 Bootstrapping is a kind of sampling algorithm. The basic concept of bootstrapping is to control the appearance time of the training instances in the sampled samples.…”