arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2102.07106stat.MLcs.LG
收藏

高斯过程的愈合产物

Healing Products of Gaussian Processes

University College London(伦敦大学学院);University of Johannesburg(约翰内斯堡大学);University of the Witwatersrand(金山大学)

Samuel Cohen, Rendani Mbuvha, Tshilidzi Marwala, Marc Peter Deisenroth

摘要:

高斯过程(GP)是一种非参数贝叶斯模型,已被应用于回归和分类问题。减轻其立方训练成本的方法之一是使用在数据子集上训练的本地GP专家。特别是,专家产品模型通过一个易于处理的产品操作结合了当地专家的预测分布。虽然这些专家模型允许大规模分布式计算,但它们的预测通常会受到平均值或未校准不确定性量化的不稳定行为的影响。通过校准预测通过回火softmax加权,我们提供了一个解决方案,这些问题的多个产品的专家模型,包括广义产品的专家和强大的贝叶斯委员会机。此外,我们利用最佳运输文献,并提出了一个新的产品的专家模型,结合当地专家的预测,通过计算他们的Wasserstein重心,这可以应用于回归和分类。

英文摘要:

Gaussian processes (GPs) are nonparametric Bayesian models that have been applied to regression and classification problems. One of the approaches to alleviate their cubic training cost is the use of local GP experts trained on subsets of the data. In particular, product-of-expert models combine the predictive distributions of local experts through a tractable product operation. While these expert models allow for massively distributed computation, their predictions typically suffer from erratic behaviour of the mean or uncalibrated uncertainty quantification. By calibrating predictions via a tempered softmax weighting, we provide a solution to these problems for multiple product-of-expert models, including the generalised product of experts and the robust Bayesian committee machine. Furthermore, we leverage the optimal transport literature and propose a new product-of-expert model that combines predictions of local experts by computing their Wasserstein barycenter, which can be applied to both regression and classification.

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。

↑