Evolutionary computation in machine learning is mainly focused on solving optimization problems that fail by using classical numerical methods with poor results. Evolutionary computation offers a reliable and effective approach to address complex problems in real-world applications in wider application domains such as agriculture, manufacturing, power and energy, networking, finance, and healthcare. Evolutionary computation(EC) is a subfield of computer intelligence, and an optimization methodology refers to the mechanisms of biological evolution and behavior of living organisms. It models the essential elements of biological evolution and explores the solution space by gene inheritance and mutation.
Evolutionary approaches in machine learning involve prepossessing, learning, and post-processing. Enhanced-EC algorithm (MLEC) stores sample data about the problem features, search space, and population information during the iterative search process. Thus the machine learning technique helps analyze these data for enhancing the search performance. EC algorithms include genetic algorithm (GA), evolutionary programming (EP), evolutionary strategies (ES), genetic programming (GP), learning classifier systems (LCS), differential evolution (DE), estimation of distribution algorithm(EDA), and memetic algorithms. Future scopes of Evolutionary computation are Combining modular evolutionary machine learning, Transfer learning using machine learning models, and Multi/many-objective evolutionary machine learning.
• Evolutionary computation (EC) is an optimization methodology inspired by the mechanisms of biological evolution and behaviors of living organisms.
• During the iterative search process, the EC algorithm stores ample data about the search space, problem features, and population information in the real-world environment.
• Machine learning in Evolutionary Computation enhances the algorithmic performance by acquiring useful information from data that is stored by EC algorithms through statistical methods ( mean and variance), interpolation and regression, clustering analysis (CA), principal component analysis (PCA), orthogonal experimental design (OED), opposition-based learning (OBL), artificial neural networks (ANN), support vector machines (SVM), case-based reasoning, reinforcement learning, competitive learning, and Bayesian network.
• Machine learning techniques are incorporated into different EC algorithms and involved in various ways and affect EC in various aspects to solve the optimization problems.
• EC algorithms are appealing for a wider range of complex real-world applications owing to the improved search speed and accuracy.
• The applications of ML-enhanced EC algorithms in multi-objective, dynamic and uncertain, large-scale, and constrained problems are still insufficient. Thus, effective and efficient MLEC algorithms become necessary to solve these complex problems in the future.