A fundamental study on the effectiveness of Immune Algorithm for Multi-objective 0/1 knapsack problem

Immune Algorithms (IAs) are categorized into three classes. Although the major type of IAs using clonal selection have been widely investigated and applied to Multi-objective Optimization Problems (MOPs), IAs based on self-regulation by suppressor T-cells have not. This paper focuses on the latter IAs and proposes Non-dominated Prioritized IA (NPIA) which is designed to solve MOPs with keeping the characteristics of IAs without clonal selection. Experimental results have shown NPIA´s search performance competitive with MOGAs and possibility of IA-based multi-objective optimization.