Applying electromagnetism-like mechanism for feature selection

Bi-objective integrating sustainable order allocation and sustainable supply chain network strategic design with stochastic demand using a novel robust hybrid multi-objective metaheuristic

“…It is a population-based metaheuristic structure to search for the optimal solution in continuous optimization problems (Tsou and Kao, 2006) and has rarely been used for discrete optimization problems (Vahdani and Zandieh, 2010). EMA utilizes the attraction-repulsion mechanism of electromagnetism theory, which is based on Coulomb's law to determine the optimal solution (Su and Lin, 2011). EMA and evolutionary algorithms are similar to each other in terms of structural aspects; mainly, they are populationbased approaches with information sharing among members.…”

“…These particles are randomly chosen from the feasible region, which is an N dimensional hyperspace. The initial value is assumed to be uniformly distributed between the corresponding upper and lower bounds (Su and Lin, 2011). Upon the particle generation, a conversion method is used to parse the primary solutions with the corresponding values of each particle in the discrete area as illustrated in Fig.…”

Physics-Inspired Optimization Algorithms: A Survey

“…Numerous applications are developed on the basis of this algorithm such as scheduling problems [147][148][149][150], course timetabling [151], PID controller [152], fuzzy system [153][154][155], vehicle routing problem [156], networking [157], inventory control [158], neural network [159,160], TSP [161,162], feature selection [163], antenna application [164], robotics application [165], flow path designing [166], and vehicle routing [167].…”

Bare bones particle swarm optimization with adaptive chaotic jump for feature selection in classification

“…By discarding irrelevant or redundant features, a smaller feature subset has three main advantages: 1) reduce the computational cost; 2) improve the performance of classifier and avoid over-fitting; 3) enhance the interpretation ability of the classification model. According to the method used to evaluate the feature subsets, FS approaches can be divided into two categories: wrapper 3,4 and filter 5 approaches. The wrapper approach uses a given learning algorithm to evaluate the feature subsets while the filter approach utilizes the inherent characteristics of the dataset to evaluate the feature subsets, such as the correlation, redundancy, and statistical dependence 6,7 .…”