China
Africa
Explore chapters and articles related to this topic
Genetic Algorithm (GA)
Published in Paresh Chra Deka, A Primer on Machine Learning Applications in Civil Engineering, 2019
The ability to search the solution space and locate regions that potentially contain optimal solutions for a given problem is one of the fundamental components of most artificial intelligence (AI) systems. There are three primary types of search: the blind search, hill climbing, and beam search. GP is classified as a beam search because it maintains a population of solutions that is smaller than all of the available solutions. GP is also usually implemented as a weak search algorithm as it contains no problem-specific knowledge, although some research has been directed towards ‘strongly typed genetic programming.’ However, while GP can find regions containing optimal solutions, an additional local search algorithm is normally required to locate the optima. Memetic algorithms can fulfill this role by combining an evolutionary algorithm with problem-specific search algorithm to locate optimal solutions.
Cyber-attack feature processing approach based on MA-LSSVM
Published in Xiaoling Jia, Feng Wu, Electromechanical Control Technology and Transportation, 2017
Moscato P (1989) proposed the memetic algorithm for the first time. In the standard genetic algorithm process, optimal solution was obtained generally through individual mutation, crossover, and selection as well as adaptive iterative evolution to the individual (Zheng Yamin, 2008). However, for the memetic algorithm, genetic operation object is not ordinary individuals in population space, but the excellent individuals elected in each local area. Genetic operation is to select the adaptable outstanding individual. In addition, new individuals can be generated through crossover operation. Such new individuals may belong to some new areas and will be replaced by adjacent superior individuals in the next generation of local search. And then perform a new round of global evolution. The search solution efficiency of the memetic algorithm based on the global and local search mechanisms is several orders of magnitude higher than that of the traditional genetic algorithms on certain issues. It can be applied to a wide range of areas and satisfactory results can be obtained. For example, Karaoglan I (2015) applied the memetic algorithm to solve mixed cargo. Cattaruzza D (2014) applied the memetic algorithm on studies of vehicle path planning.
Introduction
Published in Joseph Y.-T. Leung, Handbook of SCHEDULING, 2004
We have explored various algorithmic variants of combining TS and evolutionary methods (see [20] for more details). We have considered the following approaches: Multistart TS with random planning order. It enables a direct comparison between TS, memetic algorithms and hybrids. Finally, the best solution is improved with the greedy shuffle.Memetic algorithm with improved local search. This option applies the TS2 algorithm from Section 44.5.2 to the best solution obtained by the memetic algorithm that copies the best 4 assignments described in Section 44.5.3.1. The value x=4 turned out to produce the best results for a wide range of problems.Switch. All the previously introduced algorithms satisfy the user determined hard constraints but this approach accepts all solutions that remain within the minimum-preferred requirements interval. New generations are created by adding or removing assignments (within the feasible region) in the parent schedules.
An Effective Chronic Disease Prediction using Multi-Objective Firefly Optimisation Random Forest Algorithm
Published in IETE Journal of Research, 2022
Over the past few decades, researchers experienced that by applying metaheuristic algorithms, the model's efficiency has improved [10] because of the potential to discover optimal solutions quickly. Metaheuristic algorithms make use of heuristic knowledge to explore the optimal solution in a reasonable period. This kind of algorithm is majorly classified into single solution-based and population-based algorithm [11]. The most commonly used single solution-based algorithms are simulated annealing and Tabu search [12,13]. Some prevalent population-based algorithms are genetic algorithm, particle swarm optimisation, and ant colony optimisation [14–16]. Recently, another type of optimisation algorithm joined with this class is a memetic algorithm (MA). This algorithm is a hybrid method that composed of an evolutionary algorithm and a general optimisation algorithm [17].
The unequal area facility layout problem with shortest single-loop AGV path: how material handling method matters
Published in International Journal of Production Research, 2021
Amir Ahmadi-Javid, Amir Ardestani-Jaafari
In the literature of FLPs, evolutionary methods are the most commonly used metaheuristics (local-search methods such as Tabu Search (TS) and SA are also applied). Genetic algorithms are the most popular among evolutionary algorithms, and they are known to be very effective for unequal area FLPs, as demonstrated in recent studies by Gonçalves and Resende (2015), Paes, Artur Pessoa, and Vidal (2017), and Palomo-Romero, Salas-Morera, and García-Hernández (2017). The main distinguishing characteristics of Genetic algorithms are creating new solutions (children) by recombination (crossover) of existing solutions (parents) and avoiding convergence to a local optimum by perturbation (mutation). Genetic algorithms hybridised with local search methods are usually referred to as Memetic algorithms.
A fuzzy irregular cellular automata-based method for the vertex colouring problem
Published in Connection Science, 2020
Mostafa Kashani, Saeid Gorgin, Seyed Vahab Shojaedini
Memetic algorithms, also known as genetic local search, are formed from combining the evolutionary algorithms and local search methods. In this set of methods, the evolution and local improvement of the individuals are conducted separately. This group of methods has also been widely used for GC. Lü and Hao (2010) introduced a memetic algorithm called MACOL for GC. In this method, they have used a genetic algorithm with a new adaptive multi-parent crossover for generating offspring. In addition, they have used a Tabu search operator to improve the quality of the initial population of the algorithm. Fister, Yang, Fister, and Brest (2012) proposed a memetic firefly algorithm for solving the three-colouring graph problem.