arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1906.08324cs.LGstat.ML
收藏

功能性神经过程

The Functional Neural Process

University of Amsterdam(阿姆斯特丹大学);TNO Intelligent Imaging(TNO智能成像);Bosch Center for Artificial Intelligence(博世人工智能中心);UvA-Bosch Delta Lab(阿姆斯特丹大学-博世Delta实验室);Qualcomm(高通)

Christos Louizos, Xiahan Shi, Klamer Schutte, Max Welling

摘要:

本文提出了一类新的可交换随机过程,即功能神经过程。FNP通过在给定数据集中的点的潜在表示之上学习依赖关系图来对函数上的分布进行建模。在这样做时,他们定义了一个贝叶斯模型,而没有显式地设定潜在全局参数的先验分布;相反,他们采用给定数据集的关系结构的先验,这是一项简单得多的任务。我们展示了如何从数据中学习这些模型,证明它们可以通过小批量优化扩展到大型数据集,并描述了如何通过后验预测分布对新点进行预测。我们实验评估FNP的玩具回归和图像分类的任务,并表明,相比基线,采用全球潜在参数,他们提供了竞争力的预测,以及更强大的不确定性估计。

英文摘要:

We present a new family of exchangeable stochastic processes, the Functional Neural Processes (FNPs). FNPs model distributions over functions by learning a graph of dependencies on top of latent representations of the points in the given dataset. In doing so, they define a Bayesian model without explicitly positing a prior distribution over latent global parameters; they instead adopt priors over the relational structure of the given dataset, a task that is much simpler. We show how we can learn such models from data, demonstrate that they are scalable to large datasets through mini-batch optimization and describe how we can make predictions for new points via their posterior predictive distribution. We experimentally evaluate FNPs on the tasks of toy regression and image classification and show that, when compared to baselines that employ global latent parameters, they offer both competitive predictions as well as more robust uncertainty estimates.

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。

↑