International Journal of Electrical and Computer Engineering (IJECE) Vol. 13, No. 6, December 2023, pp. 7016~7026 ISSN: 2088-8708, DOI: 10.11591/ijece.v13i6.pp7016-7026  7016 Journal homepage: http://ijece.iaescore.com Best-worst northern goshawk optimizer: a new stochastic optimization method Purba Daru Kusuma, Faisal Candrasyah Hasibuan Department of Computer Engineering, Faculty of Electrical Engineering, Telkom University, Bandung, Indonesia Article Info ABSTRACT Article history: Received Apr 12, 2023 Revised Jul 14, 2023 Accepted Jul 17, 2023 This study introduces a new metaheuristic method: the best-worst northern goshawk optimizer (BW-NGO). This algorithm is an enhanced version of the northern goshawk optimizer (NGO). Every BW-NGO iteration consists of four phases. First, each agent advances toward the best agent and away from the worst agent. Second, each agent moves relatively to the agent selected at random. Third, each agent conducts a local search. Fourth, each agent traces the space at random. The first three phases are mandatory, while the fourth phase is optional. Simulation is performed to assess the performance of BW-NGO. In this simulation, BW-NGO is confronted with four algorithms: particle swarm optimization (PSO), pelican optimization algorithm (POA), golden search optimizer (GSO), and northern goshawk optimizer (NGO). The result exhibits that BW-NGO discovers an acceptable solution for the 23 benchmark functions. BW-NGO is better than PSO, POA, GSO, and NGO in consecutively optimizing 22, 20, 15, and 11 functions. BW-NGO can discover the global optimal solution for three functions. Keywords: Agent system Local search Metaheuristic Northern goshawk optimization Particle swarm optimization Stochastic optimization Swarm intelligence This is an open access article under the CC BY-SA license. Corresponding Author: Purba Daru Kusuma Department of Computer Engineering, Faculty of Electrical Engineering, Telkom University Bandung, Indonesia Email: purbodaru@telkomuniversity.ac.id 1. INTRODUCTION Metaheuristic is a technique that is extensively used in studies regarding optimization. Its popularity mainly comes from its flexibility in tackling various optimization problems, from simple ones to very complicated ones. Metaheuristic becomes flexible because it adopts an approximate approach that does not trace all possible solutions [1]. This strategy makes metaheuristics adaptive enough to solve optimization problems using limited computational resources. However, this benefit is offset by the metaheuristic choosing a suboptimal answer as the acceptable option and cannot guarantee the global optimal solution [1]. Recently, many metaheuristics have been ready to use for any optimization problems. Many old algorithms are still popular. Genetic algorithm (GA) is still implemented in recent engineering studies, such as hydrogen liquefaction process [2], town logistics distribution systems [3], construction projects [4], spectrometry peak detection [5], plain text encryption [6], Parkinson disease prediction [7], and so on. GA is also utilized in optimization studies related to the financial sector, such as in-stock selection [8], credit rating assessment [9], and cryptocurrency price forecasting [10]. Artificial bee colony (ABC) is also used in a lot of optimization studies in the engineering sector, such as in power contract capacity optimization [11], parallel machine scheduling [12], tunnel deformation prediction [13], and hybrid flow-shop scheduling [14]. Besides these old algorithms, many later algorithms have received positive attention. Grey wolf optimizer (GWO) has been modified and implemented in several engineering studies, such as in flood evacuation planning [15],