Pretext Training Algorithms for Event Sequence Data
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866916127969378304 |
|---|---|
| author | Wang, Yimu Zhao, He Deng, Ruizhi Tung, Frederick Mori, Greg |
| author_facet | Wang, Yimu Zhao, He Deng, Ruizhi Tung, Frederick Mori, Greg |
| contents | Pretext training followed by task-specific fine-tuning has been a successful approach in vision and language domains. This paper proposes a self-supervised pretext training framework tailored to event sequence data. We introduce a novel alignment verification task that is specialized to event sequences, building on good practices in masked reconstruction and contrastive learning. Our pretext tasks unlock foundational representations that are generalizable across different down-stream tasks, including next-event prediction for temporal point process models, event sequence classification, and missing event interpolation. Experiments on popular public benchmarks demonstrate the potential of the proposed method across different tasks and data domains. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2402_10392 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Pretext Training Algorithms for Event Sequence Data Wang, Yimu Zhao, He Deng, Ruizhi Tung, Frederick Mori, Greg Machine Learning Artificial Intelligence Pretext training followed by task-specific fine-tuning has been a successful approach in vision and language domains. This paper proposes a self-supervised pretext training framework tailored to event sequence data. We introduce a novel alignment verification task that is specialized to event sequences, building on good practices in masked reconstruction and contrastive learning. Our pretext tasks unlock foundational representations that are generalizable across different down-stream tasks, including next-event prediction for temporal point process models, event sequence classification, and missing event interpolation. Experiments on popular public benchmarks demonstrate the potential of the proposed method across different tasks and data domains. |
| title | Pretext Training Algorithms for Event Sequence Data |
| topic | Machine Learning Artificial Intelligence |
| url | https://arxiv.org/abs/2402.10392 |