Knapsack Problem With Cardinality Constraint: A Faster FPTAS Through the Lens of Numerical Analysis and Applications
share
We study the K-item knapsack problem (, 1.5-dimensional knapsack problem), which is a generalization of the famous 0-1 knapsack problem (, 1-dimensional knapsack problem) in which an upper bound K is imposed on the number of items selected. This problem is of fundamental importance and is known to have a broad range of applications in various fields such as computer science and operation research. It is well known that, there is no FPTAS for the d-dimensional knapsack problem when d≥ 2, unless P = NP. While the K-item knapsack problem is known to admit an FPTAS, the complexity of all existing FPTASs have a high dependency on the cardinality bound K and approximation error ε, which could result in inefficiencies especially when K and ε^-1 increase. The current best results are due to mastrolilli2006hybrid, in which two schemes are presented exhibiting a space-time tradeoff--one scheme with time complexity O(n+Kz^2/ε^2) and space complexity O(n+z^3/ε), while another scheme requires a run-time of O(n+(Kz^2+z^4)/ε^2) but only needs O(n+z^2/ε) space, where z={K,1/ε}. In this paper we close the space-time tradeoff exhibited in mastrolilli2006hybrid by designing a new FPTAS with a running time of O(n)+O(z^2·{K,ε^-2}), while simultaneously reaching the O(n+z^2/ε) space complexity. Our scheme provides O({K,ε^-2}) and O(z) improvements on the long-established state-of-the-art algorithms in time and space complexity respectively. An salient feature of our algorithm is that it is the first FPTAS, which achieves better time and space complexity bounds than the very first standard FPTAS over all parameter regimes.
READ FULL TEXT