Details
| Originalsprache | Englisch |
|---|---|
| Seiten (von - bis) | 16023-16031 |
| Seitenumfang | 9 |
| Fachzeitschrift | Proceedings of the AAAI Conference on Artificial Intelligence |
| Jahrgang | 38 |
| Ausgabenummer | 14 |
| Publikationsstatus | Veröffentlicht - 24 März 2024 |
| Veranstaltung | 38th AAAI Conference on Artificial Intelligence, AAAI 2024 - Vancouver, Kanada Dauer: 20 Feb. 2024 → 27 Feb. 2024 |
Abstract
Continuous-time models such as Neural ODEs and Neural Flows have shown promising results in analyzing irregularly sampled time series frequently encountered in electronic health records. Based on these models, time series are typically processed with a hybrid of an initial value problem (IVP) solver and a recurrent neural network within the variational autoencoder architecture. Sequentially solving IVPs makes such models computationally less efficient. In this paper, we propose to model time series purely with continuous processes whose state evolution can be approximated directly by IVPs. This eliminates the need for recurrent computation and enables multiple states to evolve in parallel. We further fuse the encoder and decoder with one IVP solver utilizing its invertibility, which leads to fewer parameters and faster convergence. Experiments on three real-world datasets show that the proposed method can systematically outperform its predecessors, achieve state-of-the-art results, and have significant advantages in terms of data efficiency.
ASJC Scopus Sachgebiete
- Informatik (insg.)
- Artificial intelligence
Zitieren
- Standard
- Harvard
- Apa
- Vancouver
- BibTex
- RIS
in: Proceedings of the AAAI Conference on Artificial Intelligence, Jahrgang 38, Nr. 14, 24.03.2024, S. 16023-16031.
Publikation: Beitrag in Fachzeitschrift › Konferenzaufsatz in Fachzeitschrift › Forschung › Peer-Review
}
TY - JOUR
T1 - IVP-VAE
T2 - 38th AAAI Conference on Artificial Intelligence, AAAI 2024
AU - Xiao, Jingge
AU - Basso, Leonie
AU - Nejdl, Wolfgang
AU - Ganguly, Niloy
AU - Sikdar, Sandipan
N1 - Funding Information: This research was funded by the Federal Ministry of Education and Research (BMBF), Germany under the project LeibnizKILabor with grant No. 01DD20003
PY - 2024/3/24
Y1 - 2024/3/24
N2 - Continuous-time models such as Neural ODEs and Neural Flows have shown promising results in analyzing irregularly sampled time series frequently encountered in electronic health records. Based on these models, time series are typically processed with a hybrid of an initial value problem (IVP) solver and a recurrent neural network within the variational autoencoder architecture. Sequentially solving IVPs makes such models computationally less efficient. In this paper, we propose to model time series purely with continuous processes whose state evolution can be approximated directly by IVPs. This eliminates the need for recurrent computation and enables multiple states to evolve in parallel. We further fuse the encoder and decoder with one IVP solver utilizing its invertibility, which leads to fewer parameters and faster convergence. Experiments on three real-world datasets show that the proposed method can systematically outperform its predecessors, achieve state-of-the-art results, and have significant advantages in terms of data efficiency.
AB - Continuous-time models such as Neural ODEs and Neural Flows have shown promising results in analyzing irregularly sampled time series frequently encountered in electronic health records. Based on these models, time series are typically processed with a hybrid of an initial value problem (IVP) solver and a recurrent neural network within the variational autoencoder architecture. Sequentially solving IVPs makes such models computationally less efficient. In this paper, we propose to model time series purely with continuous processes whose state evolution can be approximated directly by IVPs. This eliminates the need for recurrent computation and enables multiple states to evolve in parallel. We further fuse the encoder and decoder with one IVP solver utilizing its invertibility, which leads to fewer parameters and faster convergence. Experiments on three real-world datasets show that the proposed method can systematically outperform its predecessors, achieve state-of-the-art results, and have significant advantages in terms of data efficiency.
UR - http://www.scopus.com/inward/record.url?scp=85189645393&partnerID=8YFLogxK
U2 - 10.48550/arXiv.2305.06741
DO - 10.48550/arXiv.2305.06741
M3 - Conference article
AN - SCOPUS:85189645393
VL - 38
SP - 16023
EP - 16031
JO - Proceedings of the AAAI Conference on Artificial Intelligence
JF - Proceedings of the AAAI Conference on Artificial Intelligence
SN - 2159-5399
IS - 14
Y2 - 20 February 2024 through 27 February 2024
ER -