Using genetic algorithms for batch selection decisions

摘要

Combinatorially complex decisions are usually characterized by their huge size of solution space. For such type of decisions, optimal solutions are usually unattainable, and we can only approximate the optimal solutions. Due to their determinism, most of the traditional optimization techniques are limited by their power to discover satisfactory solutions for combinatorially complex decisions. In this paper, we designed a genetic algorithm-based interactive system for a combinatorially complex decision—the batch selection problem. A performance comparison with a rule-based system was also conducted. We also investigated the influence of different parameter values on the system performance.