Evaluating the Local Ratio Algorithm for Dynamic Storage Allocation
We empirically compare the local ratio algorithm for the profit maximization version of the dynamic storage allocation problem against various greedy algorithms. Our main conclusion is that, at least on our input distributions, the local ratio algorithms performed worse on average than the more naive greedy algorithms.
KeywordsGreedy Algorithm Competitive Ratio Experimental Space Input Distribution Horizontal Slice
Unable to display preview. Download preview PDF.
- 1.A. Barnoy, R. Bar-Yehuda, A. Freund, J. Naor, and B. Schieber, ”A unified approach to approximating resource allocation and scheduling”, ACM Symposium on Theory of Computing, 2000. To appear in JACM.Google Scholar
- 3.J. Gergov, “Algorithms for compile-time memory allocation”, ACM/SIAM Symposium on Discrete Algorithms, S907–S908, 1999.Google Scholar
- 5.S. Leonardi, A. Marchetti-Spaccamela and A. Vitaletti, “Approximation algorithms for bandwidth and storage allocation problems under real time constraints”, Conference on Foundations of Software Technology and Theoretical Computer Science (FSTTCS), 2000.Google Scholar
- 6.P. Wilson, M. Johnstone, M. Neely, and D. Boles, “Dynamic storage allocation: a survey and critical review”, International Workshop on Memory Management, Lecture Notes in Computer Science, 986, 1–116, 1995.Google Scholar