TY - GEN
T1 - Model Selection with a Shapelet-Based Distance Measure for Multi-source Transfer Learning in Time Series Classification
AU - Lee, Jiseok
AU - Iwana, Brian Kenji
N1 - Publisher Copyright:
© The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.
PY - 2025
Y1 - 2025
N2 - Transfer learning is a common practice that alleviates the need for extensive data to train neural networks. It is performed by pre-training a model using a source dataset and fine-tuning it for a target task. However, not every source dataset is appropriate for each target dataset, especially for time series. In this paper, we propose a novel method of selecting and using multiple datasets for transfer learning for time series classification. Specifically, our method combines multiple datasets as one source dataset for pre-training neural networks. Furthermore, for selecting multiple sources, our method measures the transferability of datasets based on shapelet discovery for effective source selection. While traditional transferability measures require considerable time for pre-training all the possible sources for source selection of each possible architecture, our method can be repeatedly used for every possible architecture with a single simple computation. Using the proposed method, we demonstrate that it is possible to increase the performance of temporal convolutional neural networks (CNN) on time series datasets.
AB - Transfer learning is a common practice that alleviates the need for extensive data to train neural networks. It is performed by pre-training a model using a source dataset and fine-tuning it for a target task. However, not every source dataset is appropriate for each target dataset, especially for time series. In this paper, we propose a novel method of selecting and using multiple datasets for transfer learning for time series classification. Specifically, our method combines multiple datasets as one source dataset for pre-training neural networks. Furthermore, for selecting multiple sources, our method measures the transferability of datasets based on shapelet discovery for effective source selection. While traditional transferability measures require considerable time for pre-training all the possible sources for source selection of each possible architecture, our method can be repeatedly used for every possible architecture with a single simple computation. Using the proposed method, we demonstrate that it is possible to increase the performance of temporal convolutional neural networks (CNN) on time series datasets.
KW - Time Series Classification
KW - Transfer Learning
KW - Transferability Estimation
UR - http://www.scopus.com/inward/record.url?scp=85211803582&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85211803582&partnerID=8YFLogxK
U2 - 10.1007/978-3-031-78398-2_11
DO - 10.1007/978-3-031-78398-2_11
M3 - Conference contribution
AN - SCOPUS:85211803582
SN - 9783031783975
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 160
EP - 175
BT - Pattern Recognition - 27th International Conference, ICPR 2024, Proceedings
A2 - Antonacopoulos, Apostolos
A2 - Chaudhuri, Subhasis
A2 - Chellappa, Rama
A2 - Liu, Cheng-Lin
A2 - Bhattacharya, Saumik
A2 - Pal, Umapada
PB - Springer Science and Business Media Deutschland GmbH
T2 - 27th International Conference on Pattern Recognition, ICPR 2024
Y2 - 1 December 2024 through 5 December 2024
ER -