ShanghaiTech University Knowledge Management System
| CNN KERNELS CAN BE THE BEST SHAPELETS | |
| 2024 | |
| 会议录名称 | 12TH INTERNATIONAL CONFERENCE ON LEARNING REPRESENTATIONS, ICLR 2024 |
| 摘要 | Shapelets and CNN are two typical approaches to model time series. Shapelets aim at finding a set of sub-sequences that extract feature-based interpretable shapes, but may suffer from accuracy and efficiency issues. CNN performs well by encoding sequences with a series of hidden representations, but lacks interpretability. In this paper, we demonstrate that shapelets are essentially equivalent to a specific type of CNN kernel with a squared norm and pooling. Based on this finding, we propose ShapeConv, an interpretable CNN layer with its kernel serving as shapelets to conduct time-series modeling tasks in both supervised and unsupervised settings. By incorporating shaping regularization, we enforce the similarity for maximum interpretability. We also find human knowledge can be easily injected to ShapeConv by adjusting its initialization and model performance is boosted with it. Experiments show that ShapeConv can achieve state-of-the-art performance on time-series benchmarks without sacrificing interpretability and controllability. © 2024 12th International Conference on Learning Representations, ICLR 2024. All rights reserved. |
| 关键词 | Benchmarking Encodings Feature-based Human knowledge Interpretability Modeling performance Modeling task Regularisation Shapelets Times series Times series models |
| 会议名称 | 12th International Conference on Learning Representations, ICLR 2024 |
| 会议地点 | Hybrid, Vienna, Austria |
| 会议日期 | May 7, 2024 - May 11, 2024 |
| 收录类别 | EI |
| 语种 | 英语 |
| 出版者 | International Conference on Learning Representations, ICLR |
| EI入藏号 | 20243216836987 |
| EI主题词 | Time series |
| EI分类号 | 922.2 Mathematical Statistics |
| 原始文献类型 | Conference article (CA) |
| 文献类型 | 会议论文 |
| 条目标识符 | https://kms.shanghaitech.edu.cn/handle/2MSLDSTB/411263 |
| 专题 | 信息科学与技术学院_PI研究组_任侃组 |
| 通讯作者 | Luo, Xufang; Li, Dongsheng |
| 作者单位 | 1.University of California, Berkeley, United States 2.Microsoft Research Asia, China 3.University of Science and Technology of China, China 4.ShanghaiTech University, China |
| 推荐引用方式 GB/T 7714 | Qu, Eric,Wang, Yansen,Luo, Xufang,et al. CNN KERNELS CAN BE THE BEST SHAPELETS[C]:International Conference on Learning Representations, ICLR,2024. |
| 条目包含的文件 | ||||||
| 条目无相关文件。 | ||||||
修改评论
除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。