NP-PROV: Neural Processes with Position-Relevant-Only Variances

Xuesong Wang, Lina Yao, Xianzhi Wang, Feiping Nie, Boualem Benatallah

科研成果: 书/报告/会议事项章节会议稿件同行评审

2 引用 (Scopus)

摘要

Neural Processes (NPs) families encode distributions over functions to a latent representation given a set of context data, and decode posterior mean and variance at unknown locations. Since mean and variance are derived from the same latent space, they may fail on out-of-domain tasks where fluctuations in function values amplify the model uncertainty. We present a new member named Neural Processes with Position-Relevant-Only Variances (NP-PROV). NP-PROV hypothesizes that a target point close to a context point has small uncertainty, regardless of the function value at that position. The resulting approach derives mean and variance from a function-value-related space and a position-related-only latent space separately. Our evaluation on synthetic and real-world datasets reveals that NP-PROV can achieve state-of-the-art likelihood while retaining a bounded variance when drifts exist in the function value.

源语言英语
主期刊名Web Information Systems Engineering - WISE 2021 - 22nd International Conference on Web Information Systems Engineering, WISE 2021, Proceedings
编辑Wenjie Zhang, Lei Zou, Zakaria Maamar, Lu Chen
出版商Springer Science and Business Media Deutschland GmbH
129-142
页数14
ISBN(印刷版)9783030908874
DOI
出版状态已出版 - 2021
活动22nd International Conference on Web Information Systems Engineering, WISE 2021 - Melbourne, 澳大利亚
期限: 26 10月 202129 10月 2021

出版系列

姓名Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
13080 LNCS
ISSN(印刷版)0302-9743
ISSN(电子版)1611-3349

会议

会议22nd International Conference on Web Information Systems Engineering, WISE 2021
国家/地区澳大利亚
Melbourne
时期26/10/2129/10/21

指纹

探究 'NP-PROV: Neural Processes with Position-Relevant-Only Variances' 的科研主题。它们共同构成独一无二的指纹。

引用此