Binary Competitive Swarm Optimizer Approaches for Feature Selection

Abstract

Feature selection is known as an NP-hard combinatorial problem in which the possible feature subsets increase exponentially with the number of features. Due to the increment of the feature size, the exhaustive search has become impractical. In addition, a feature set normally includes irrelevant, redundant, and relevant information. Therefore, in this paper, binary variants of a competitive swarm optimizer are proposed for wrapper feature selection. The proposed approaches are used to select a subset of significant features for classification purposes. The binary version introduced here is performed by employing the S-shaped and V-shaped transfer functions, which allows the search agents to move on the binary search space. The proposed approaches are tested by using 15 benchmark datasets collected from the UCI machine learning repository, and the results are compared with other conventional feature selection methods. Our results prove the capability of the proposed binary version of the competitive swarm optimizer not only in terms of high classification performance, but also low computational cost.

Naïve Bayes and linear discriminate analysis based diagnostic analytic of harmonic source identification

“…It is worth noting that NB and LDA are computed for 10 runs and the average results are recorded. For performance measurement, six popular evaluation metrics, namely accuracy, precision, geometric mean, sensitivity, specificity, and F-measure are used, and they can be defined as follows [3,4]:…”

“…Power quality signals are considered as the natural pathway to detect the harmonic sources, and since the resulting voltage and current are the additions of multiple action potential traversing the electric circuit [1,2]. Research in the identification, detection and classification using machine learning algorithm has become one of the major interests, which allows the multiple harmonic sources identification become viable [3][4][5]. In the identification process, the features are first extracted, and then the machine learning algorithm is implemented in order to generate a better understanding around the potential features or solutions [6][7][8].…”

“…The NB is a simple classifier based on Bayes' theorem that assumes all the features are independence [36], whereas the linear discriminate analysis (LDA) is faster and computationally less expensive. At last, the best machine learning method for identifying and diagnosing harmonic sources is selected based on the performance measurement criteria such as the accuracy, precision, geometric mean, and F-measure [3].…”

Hybrid Binary Particle Swarm Optimization Differential Evolution-Based Feature Selection for EMG Signals Classification

“…Then, BPSO (iteration with odd number) and BDE (iteration with even number) algorithms are computed in sequence. For the iteration with odd number, the inertia weight is updated as shown in Equation (10). Afterward, the position and velocity of particles are updated using Equations (1) and 3, respectively.…”

“…Compared to the filter approach, the wrapper approach can often contribute to a better performance. Hence, wrapper approaches are widely used in feature selection problems [6,[8][9][10].…”