Abstract
In the two-thinning balls-and-bins model, an overseer is provided with uniform random allocation of m balls into n bins in an on-line fashion. The overseer may reject the allocation of each ball, in which case it is placed into a new bin, drawn independently, uniformly at random. The purpose of the overseer is to reduce the maximum load, that is, the difference between the maximum number of balls in a single bin and the average number of balls among all bins.
We provide tight estimates for three quantities: the lowest achievable maximum load at a given time m, the lowest achievable maximum load uniformly over the entire time interval and the lowest achievable typical maximum load over the interval , that is, a load which upper-bounds portion of the times in .
In particular, for m polynomial in n and sufficiently large, we provide an explicit strategy, which achieves a typical maximum load of , asymptotically the same as that can be achieved at a single time m. In contrast, we show that no strategy can achieve better than maximum load for all times up to time m.
Funding Statement
The first author was supported by the ISF Grant 1327/19.
The second author was supported by the ISF Grant 1695/20.
Acknowledgments
The third author was with the Hebrew University of Jerusalem during the preparation of this paper. We thank the referees for their very careful reading of the paper and many valuable comments.
Citation
Ohad Noy Feldheim. Ori Gurel-Gurevich. Jiange Li. "Long-term balanced allocation via thinning." Ann. Appl. Probab. 34 (1A) 795 - 850, February 2024. https://doi.org/10.1214/23-AAP1978
Information