Towards Learning to Rank Deep-Learning Models for Multivariate Time-Series Transfer Learning

Sigl M, Meyer-Wegener K (2025)

Publication Language: English

Publication Type: Conference contribution

Publication year: 2025

Publisher: Association for Computing Machinery, Inc

City/Town: New York, NY

Pages Range: 2

Conference Proceedings Title: DEEM '25: Proceedings of the Workshop on Data Management for End-to-End Machine Learning

Event location: Berlin DE

ISBN: 9798400719240

DOI: 10.1145/3735654.3735938

Abstract

Selecting an unsuitable deep-learning (DL) model for often results in negative transfer and poor performance. To address this, we propose a novel ranking approach designed to guide DL model selection by leveraging dataset characteristics in multivariate time-series data. Our approach considers three dataset characteristics: statistical, shape-based, and a combination of both. By implicitly learning dataset similarities, the ranking model identifies the most suitable DL model for positive transfer learning. Experiments show that a ranking model effectively learns to rank the most suitable DL model at position one. The impact of characteristics on the ranking model's performance depends on the size of the training rankings. This approach highlights the importance of learning to select DL models for on multivariate time-series datasets and offers a practical solution for effective in time-series applications.

Authors with CRIS profile

Melanie Sigl Chair for Computer Science 6 (Data Management) Klaus Meyer-Wegener Technische Fakultät

How to cite

APA:

Sigl, M., & Meyer-Wegener, K. (2025). Towards Learning to Rank Deep-Learning Models for Multivariate Time-Series Transfer Learning. In DEEM '25: Proceedings of the Workshop on Data Management for End-to-End Machine Learning (pp. 2). Berlin, DE: New York, NY: Association for Computing Machinery, Inc.

MLA:

Sigl, Melanie, and Klaus Meyer-Wegener. "Towards Learning to Rank Deep-Learning Models for Multivariate Time-Series Transfer Learning." Proceedings of the 9th Workshop on Data Management for End-to-End Machine Learning, DEEM 2025 - In conjunction with the 2025 ACM SIGMOD/PODS Conference, Berlin New York, NY: Association for Computing Machinery, Inc, 2025. 2.

BibTeX: Download