In this paper, based on the concept of the NFL theorem, that there is no unique algorithm that has the best performance for all optimization problems, a new human-based metaheuristic algorithm called Language Education Optimization (LEO) is introduced, which is used to solve optimization problems. LEO is inspired by the foreign language education process in which a language teacher trains the students of language schools in the desired language skills and rules. LEO is mathematically modeled in three phases: (i) students selecting their teacher, (ii) students learning from each other, and (iii) individual practice, considering exploration in local search and exploitation in local search. The performance of LEO in optimization tasks has been challenged against fifty-two benchmark functions of a variety of unimodal, multimodal types and the CEC 2017 test suite. The optimization results show that LEO, with its acceptable ability in exploration, exploitation, and maintaining a balance between them, has efficient performance in optimization applications and solution presentation. LEO efficiency in optimization tasks is compared with ten well-known metaheuristic algorithms. Analyses of the simulation results show that LEO has effective performance in dealing with optimization tasks and is significantly superior and more competitive in combating the compared algorithms. The implementation results of the proposed approach to four engineering design problems show the effectiveness of LEO in solving real-world optimization applications.

This paper introduces a new metaheuristic algorithm called Migration Algorithm (MA), which is helpful in solving optimization problems. The fundamental inspiration of MA is the process of human migration, which aims to improve job, educational, economic, and living conditions, and so on. The mathematical modeling of the proposed MA is presented in two phases to empower the proposed approach in exploration and exploitation during the search process. In the exploration phase, the algorithm population is updated based on the simulation of choosing the migration destination among the available options. In the exploitation phase, the algorithm population is updated based on the efforts of individuals in the migration destination to adapt to the new environment and improve their conditions. MA's performance is evaluated on fifty-two standard benchmark functions consisting of unimodal and multimodal types and the CEC 2017 test suite. In addition, MA's results are compared with the performance of twelve well-known metaheuristic algorithms. The optimization results show the proposed MA approach's high ability to balance exploration and exploitation to achieve suitable solutions for optimization problems. The analysis and comparison of the simulation results show that MA has provided superior performance against competitor algorithms in most benchmark functions. Also, the implementation of MA on four engineering design problems indicates the effective capability of the proposed approach in handling optimization tasks in real-world applications.