Provably Good Solutions to the Knapsack Problem via Neural Networks of Bounded Size

Open in new window