We consider Online Minimum Bipartite Matching under the uniform metric. We show that Randomized Greedy achieves a competitive ratio equal to $(1+1/n) (H_{n+1}-1)$, which matches the lower bound. Comparing with the fact that RG achieves an optimal ratio of $\Theta(\ln n)$ for the same problem but under the adversarial order, we find that the weaker arrival assumption of random order doesn't offer any extra algorithmic advantage for RG, or make the model strictly more tractable.
翻译:我们根据统一的衡量标准来考虑在线最低两边配对。 我们发现随机性贪婪达到了相当于1+1美元(H ⁇ n+1}-1美元)的竞争性比率,这与较低的约束值相当。 与RG为同一问题但根据对抗性命令达到1美元($)的最佳比率相比,我们发现随机性订单较弱的抵达假设不会给RG带来任何额外的算法优势,或者使模型更严格地说更具有可牵动性。