A Binary Stellar Oscillation Optimizer for 0–1 Knapsack Problems
TL;DR
The study confirms that the proposed algorithm extends the applicability of SOO to discrete domains while offering a reliable and efficient approach for solving 0–1 KP instances.
Abstract
The stellar oscillation optimizer (SOO) is a novel nature‐inspired metaheuristic algorithm, which draws its inspiration from the motion of stellar objects in the universe. Despite its excellent performance in continuous optimization, the standard SOO framework faces limitations when addressing discrete optimization problems like the 0–1 knapsack problem (0–1 KP). This paper proposes a binary stellar oscillation optimizer (BinSOO) to effectively solve 0–1 KP. This paper applies the sigmoid function to convert continuous variables into discrete ones and introduces a simple yet effective greedy‐based individual repair and optimization operator to address constraint handling. The performance of the proposed method is evaluated on three benchmark datasets containing 57 instances of the 0–1 knapsack problem. Experimental results demonstrate that BinSOO achieves highly competitive solution quality and robustness, frequently matching or surpassing several state‐of‐the‐art binary metaheuristics. The study confirms that the proposed algorithm extends the applicability of SOO to discrete domains while offering a reliable and efficient approach for solving 0–1 KP instances.