Novel quantum-inspired evolutionary algorithms for the quadratic knapsack problem
C. Patvardhan, Ved Prakash, A. Srivastav · International Journal of Mathematics in Operational Research · 2012
The Knapsack Problem assigns weights and values to each of a given set of objects, with the objective of choosing a subset no larger than some specified weight constraint in such a manner as to maximise the total value accrued thereby. The Quadratic Kapsack Problem (QKP) extends KP by associating values with not just individual objects but also with pairs of objects. The problem is known to be NP-hard and arises in several domains including finance, VLSI design and location problems. Greedy heuristics and Genetic Algorithms (GAs) for QKP exist in the literature. Quantum-inspired Evolutionary Algorithms (QEAs) are population-based probabilistic EAs that integrate concepts from quantum computing for higher representational power and robust search. This paper presents novel Quantum-inspired Evolutionary Algorithms (QEAs) that integrate concepts from quantum computing for higher representational power and robust search. The performance of the QEAs is shown to be competitive vis-à-vis two recent well known GAs over a wide range of benchmark instances.