arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 1803.09460stat.COstat.MEstat.ML
收藏

交叉随机效应模型的可扩展推断

Scalable inference for crossed random effects models

Omiros Papaspiliopoulos, Gareth O. Roberts, Giacomo Zanella

摘要:

我们分析了吉布斯采样器的复杂性,用于现代方差分析中使用的交叉随机效应模型的推理。我们证明,对于某些设计的普通香草吉布斯采样器是不可扩展的,在这个意义上说,它的复杂性比成比例的参数和数据的数量。因此,我们提出了一个简单的修改,导致一个崩溃的吉布斯采样器,可证明是可扩展的。虽然我们的理论需要对数据设计进行一些平衡性假设,但我们在模拟和真实数据集中证明,在违反假设的情况下,它预测的速率与正确的速率显著匹配。我们还表明,折叠吉布斯采样器,扩展到采样进一步未知的超参数,性能显着优于其他国家的最先进的算法。

英文摘要:

We analyze the complexity of Gibbs samplers for inference in crossed random effect models used in modern analysis of variance. We demonstrate that for certain designs the plain vanilla Gibbs sampler is not scalable, in the sense that its complexity is worse than proportional to the number of parameters and data. We thus propose a simple modification leading to a collapsed Gibbs sampler that is provably scalable. Although our theory requires some balancedness assumptions on the data designs, we demonstrate in simulated and real datasets that the rates it predicts match remarkably the correct rates in cases where the assumptions are violated. We also show that the collapsed Gibbs sampler, extended to sample further unknown hyperparameters, outperforms significantly alternative state of the art algorithms.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑