Conditional Temporal Neural Processes with Covariance Loss

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Yoo, Boseon, Lee, Jiwoo, Ju, Janghoon, Chung, Seijun, Kim, Soyeon, Choi, Jaesik
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866908294081150976
author Yoo, Boseon
Lee, Jiwoo
Ju, Janghoon
Chung, Seijun
Kim, Soyeon
Choi, Jaesik
author_facet Yoo, Boseon
Lee, Jiwoo
Ju, Janghoon
Chung, Seijun
Kim, Soyeon
Choi, Jaesik
contents We introduce a novel loss function, Covariance Loss, which is conceptually equivalent to conditional neural processes and has a form of regularization so that is applicable to many kinds of neural networks. With the proposed loss, mappings from input variables to target variables are highly affected by dependencies of target variables as well as mean activation and mean dependencies of input and target variables. This nature enables the resulting neural networks to become more robust to noisy observations and recapture missing dependencies from prior information. In order to show the validity of the proposed loss, we conduct extensive sets of experiments on real-world datasets with state-of-the-art models and discuss the benefits and drawbacks of the proposed Covariance Loss.
format Preprint
id arxiv_https___arxiv_org_abs_2504_00794
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Conditional Temporal Neural Processes with Covariance Loss
Yoo, Boseon
Lee, Jiwoo
Ju, Janghoon
Chung, Seijun
Kim, Soyeon
Choi, Jaesik
Machine Learning
Artificial Intelligence
68T07
I.2.8
We introduce a novel loss function, Covariance Loss, which is conceptually equivalent to conditional neural processes and has a form of regularization so that is applicable to many kinds of neural networks. With the proposed loss, mappings from input variables to target variables are highly affected by dependencies of target variables as well as mean activation and mean dependencies of input and target variables. This nature enables the resulting neural networks to become more robust to noisy observations and recapture missing dependencies from prior information. In order to show the validity of the proposed loss, we conduct extensive sets of experiments on real-world datasets with state-of-the-art models and discuss the benefits and drawbacks of the proposed Covariance Loss.
title Conditional Temporal Neural Processes with Covariance Loss
topic Machine Learning
Artificial Intelligence
68T07
I.2.8
url https://arxiv.org/abs/2504.00794