Parallel approximate solution of the 0/1 knapsack optimization on competitive neural networks

Summary form only given, as follows. A neural network solution to the 0/1 knapsack optimization problem is proposed. Good approximate solutions are obtained using this neural network. The neural network for 0/1 knapsack optimisation is competitive in the sense that neurons compete to become active. However, more than one neuron is allowed to be active. This is a form of controlled competition.<>