Cooperative co-evolution for feature selection in Big Data with random feature grouping

Abstract

A massive amount of data is generated with the evolution of modern technologies. This high-throughput data generation results in Big Data, which consist of many features (attributes). However, irrelevant features may degrade the classification performance of machine learning (ML) algorithms. Feature selection (FS) is a technique used to select a subset of relevant features that represent the dataset. Evolutionary algorithms (EAs) are widely used search strategies in this domain. A variant of EAs, called cooperative co-evolution (CC), which uses a divide-and-conquer approach, is a good choice for optimization problems. The existing solutions have poor performance because of some limitations, such as not considering feature interactions, dealing with only an even number of features, and decomposing the dataset statically. In this paper, a novel random feature grouping (RFG) has been introduced with its three variants to dynamically decompose Big Data datasets and to ensure the probability of grouping interacting features into the same subcomponent. RFG can be used in CC-based FS processes, hence called Cooperative Co-Evolutionary-Based Feature Selection with Random Feature Grouping (CCFSRFG). Experiment analysis was performed using six widely used ML classifiers on seven different datasets from the UCI ML repository and Princeton University Genomics repository with and without FS. The experimental results indicate that in most cases [i.e., with naïve Bayes (NB), support vector machine (SVM), k-Nearest Neighbor (k-NN), J48, and random forest (RF)] the proposed CCFSRFG-1 outperforms an existing solution (a CC-based FS, called CCEAFS) and CCFSRFG-2, and also when using all features in terms of accuracy, sensitivity, and specificity.

Optimized hybrid investigative based dimensionality reduction methods for malaria vector using KNN classifier

“…A lot of dimensionality reduction and classification approaches have been explored in literature, they are based of several measures, such as computational complexity, accuracy, among others [24]. Some of the few works done have been studies in the literature and have indicated emerging research fields such as optimization and hybridization investigations [25].…”

Infrequent Pattern Detection for Reliable Network Traffic Analysis Using Robust Evolutionary Computation

“…However, the widely used search strategy in the FS process is EC. A taxonomy of evolutionary FS approaches [2,4,6] is illustrated in Figure 3. Evolutionary FS approaches can be categorized into three types: (1) evaluation criteria-based, (2) evolutionary computation-based, (3) the number of objectives-based.…”

“…The different EC algorithms that are used in the FS process are evolutionary algorithm (EA), co-evolutionary algorithm (CEA), swarm optimization, hybrid, and other algorithms. The standard algorithms in these categories are genetic algorithm (GA), genetic programming (GP), parallel GA, cooperative co-evolutionary algorithm (CCEA), particle swarm optimization (PSO), ant colony optimization (ACO), minimum redundancy maximum relevance (mRMR), teaching learning-based algorithm (TLBO), TLBO with opposition-based learning (TLBOL), conditional mutual information maximization (CMIM), binary genetic algorithm (BGA), gravitational search algorithm (GSA), artificial bee colony (ABC), memetic algorithm (MA), and differential evolution (DE) [2,4,6].…”

“…Hence, removing unnecessary features in cybersecurity datasets is an essential preprocessing step before the main data analysis task. In addition, FS can also reduce the storage requirements [2,6]. Data fusion usually refers to integrating information gathered from different sources, for example, sensors, IoT, databases, etc.…”

“…The evolutionary computation (EC)-based FS approaches are widely used as a search technique for the FS process. However, when the search space increases because of the Big Data characteristics, traditional EC-based approaches are not suitable [2,4,6]. Recently, cooperative co-evolution (CC), a meta-heuristic-based FS approach, has shown its effectiveness in Big Data Analytics.…”