Meta-heuristics and Artificial Intelligence

Abstract

Meta-heuristics are generic search methods that are used to solve challenging combinatorial problems. We describe these methods and highlight their common features and differences by grouping them in two main kinds of approaches: Perturbative meta-heuristics that build new combinations by modifying existing combinations (such as, for example, genetic algorithms and local search), and Constructive meta-heuristics that generate new combinations in an incremental way by using a stochastic model (such as, for example, estimation of distribution algorithms and ant colony optimization). These approaches may be hybridised, and we describe some classical hybrid schemes. We also introduce the notions of diversification (exploration) and intensification (exploitation), which are shared by all these metaheuristics: diversification aims at ensuring a good sampling of the search space and, therefore, at reducing the risk of ignoring a promising sub-area which actually contains high-quality solutions, whereas intensification aims at locating the best combinations within a limited region of the search space. Finally, we describe two applications of meta-heuristics to typical artificial intelligence problems: satisfiability of Boolean formulas, and constraint satisfaction problems.

Dipper Throated Optimization Algorithm for Unconstrained Function and Feature Selection

“…Meta-heuristics refers to generic methods that normally used to solve complex and challenging combinatorial search problems. Generally, the problems solved by metaheuristic algorithms are challenging for computer scientists due to the need to examining a huge number of combinations that usually exponential with conflicting objectives [14]. Many metaheuristic algorithms have been proposed to tackle real-world situation such as image segmentation [15], water allocation and crop planning [16], Nurse Rostering [17], power load dispatch [18], and Parkinson diagnosis [19].…”

Initialization of metaheuristics: comprehensive review, critical analysis, and research directions

“…In general, local search algorithms have a good intensification capacity, while their drawbacks are mainly due to their lack of capacity to diversify the search. In fact, local searches typically focus on the exploration of the neighbors of the previously visited solutions, and therefore the exploration of new regions requires several iterations and needs the handling of different local optima (Hao and Solnon, 2020). Nevertheless, the algorithms presented in Section 2.3 provide mechanisms to maintain the diversity of the search.…”

Deployment of FACTS Controllers in Solving Power System Network Problems: A Review

“…Techniques in this category are stochastic and optimization algorithms that are based on population with efficient capability in handling multi-objective, multimodal and constrained discrete problems [30], [37]. Meta-heuristic algorithms are described as generic and search method capable of solving, combinational, complex and challenging problems [38]. In this group are; Fuzzy logic (including FCN, CWFR), genetic algorithm (GA), non-dominating sorting genetic algorithm two (NDSGA-II), artificial neural network (ANN), general regression ANN (GRANN), plant growth simulation algorithm (PGSA), body immune algorithm (BIA), particle swarm optimization (PSO), biogeography based method (BBO), modified PSO, discrete PSO, phasor discrete PSO, whale optimization (WOA) ant colony search (ACS), artificial bee colony (ABC), gradient search (GS), modified ABC etc.…”