Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
194 tokens/sec
GPT-4o
7 tokens/sec
Gemini 2.5 Pro Pro
45 tokens/sec
o3 Pro
4 tokens/sec
GPT-4.1 Pro
38 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

Multivariate Time Series Early Classification Across Channel and Time Dimensions (2306.14606v1)

Published 26 Jun 2023 in cs.LG and cs.AI

Abstract: Nowadays, the deployment of deep learning models on edge devices for addressing real-world classification problems is becoming more prevalent. Moreover, there is a growing popularity in the approach of early classification, a technique that involves classifying the input data after observing only an early portion of it, aiming to achieve reduced communication and computation requirements, which are crucial parameters in edge intelligence environments. While early classification in the field of time series analysis has been broadly researched, existing solutions for multivariate time series problems primarily focus on early classification along the temporal dimension, treating the multiple input channels in a collective manner. In this study, we propose a more flexible early classification pipeline that offers a more granular consideration of input channels and extends the early classification paradigm to the channel dimension. To implement this method, we utilize reinforcement learning techniques and introduce constraints to ensure the feasibility and practicality of our objective. To validate its effectiveness, we conduct experiments using synthetic data and we also evaluate its performance on real datasets. The comprehensive results from our experiments demonstrate that, for multiple datasets, our method can enhance the early classification paradigm by achieving improved accuracy for equal input utilization.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (63)
  1. J. Chen and X. Ran, “Deep Learning With Edge Computing: A Review,” Proceedings of the IEEE, vol. 107, no. 8, pp. 1655–1674, 2019.
  2. Z. Zhou, X. Chen, E. Li, L. Zeng, K. Luo, and J. Zhang, “Edge Intelligence: Paving the Last Mile of Artificial Intelligence With Edge Computing,” Proceedings of the IEEE, vol. 107, no. 8, pp. 1738–1762, 2019.
  3. X. Wang, Y. Han, V. C. M. Leung, D. Niyato, X. Yan, and X. Chen, “Convergence of Edge Computing and Deep Learning: A Comprehensive Survey,” IEEE Communications Surveys & Tutorials, vol. 22, no. 2, pp. 869–904, 2020.
  4. C. Canel, T. Kim, G. Zhou, C. Li, H. Lim, D. G. Andersen, M. Kaminsky, and S. Dulloor, “Scaling video analytics on constrained edge nodes,” in Proceedings of Machine Learning and Systems, 2019.
  5. J. Wang, Z. Feng, Z. Chen, S. George, M. Bala, P. Pillai, S.-W. Yang, and M. Satyanarayanan, “Bandwidth-efficient live video analytics for drones via edge computing,” in 2018 IEEE/ACM Symposium on Edge Computing (SEC).   IEEE, 2018, pp. 159–173.
  6. V. Nigade, L. Wang, and H. E. Bal, “Clownfish: Edge and cloud symbiosis for video stream analytics,” in 5th IEEE/ACM Symposium on Edge Computing, SEC, 2020, pp. 55–69.
  7. Y. Han, G. Huang, S. Song, L. Yang, H. Wang, and Y. Wang, “Dynamic neural networks: A survey,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 44, no. 11, pp. 7436–7456, 2021.
  8. Z. Huang, Z. Ye, S. Li, and R. Pan, “Length adaptive recurrent model for text classification,” in Proceedings of the 26th CIKM.   ACM, 2017, pp. 1019–1027.
  9. H. Fan, Z. Xu, L. Zhu, C. Yan, J. Ge, and Y. Yang, “Watching a small portion could be as good as watching all: Towards efficient video classification,” in Proceedings of the 27th IJCAI, 2018, pp. 705–711.
  10. A. Gupta, H. P. Gupta, B. Biswas, and T. Dutta, “Approaches and Applications of Early Classification of Time Series: A Review,” IEEE Transactions on Artificial Intelligence, vol. 1, no. 1, pp. 47–61, 2020.
  11. R. Wu, A. Der, and E. J. Keogh, “When is Early Classification of Time Series Meaningful?” IEEE Transactions on Knowledge and Data Engineering, vol. 35, no. 3, pp. 3253–3260, 2023.
  12. Y. Achenchabe, A. Bondu, A. Cornuéjols, and V. Lemaire, “Early Classification of Time Series is Meaningful,” CoRR, vol. abs/2104.13257, 2021.
  13. D. K. Dennis, C. Pabbaraju, H. V. Simhadri, and P. Jain, “Multiple Instance Learning for Efficient Sequential Data Classification on Resource-constrained Devices,” in NeurIPS, 2018, pp. 10 976–10 987.
  14. M. Rußwurm, S. Lefèvre, N. Courty, R. Emonet, M. Körner, and R. Tavenard, “End-to-end Learning for Early Classification of Time Series,” CoRR, vol. abs/1901.10681, 2019.
  15. T. Kannan and H. Hoffmann, “Budget RNNs: Multi-Capacity Neural Networks to Improve In-Sensor Inference Under Energy Budgets,” in 27th IEEE Real-Time and Embedded Technology and Applications Symposium, RTAS, 2021, pp. 143–156.
  16. J. Koutník, K. Greff, F. J. Gomez, and J. Schmidhuber, “A clockwork RNN,” in Proceedings of the 31th ICML, ser. JMLR Workshop and Conference Proceedings, vol. 32, 2014, pp. 1863–1871.
  17. V. Campos, B. Jou, X. Giró-i-Nieto, J. Torres, and S. Chang, “Skip RNN: Learning to Skip State Updates in Recurrent Neural Networks,” in 6th International Conference on Learning Representations, ICLR, 2018.
  18. G. Gobieski, B. Lucia, and N. Beckmann, “Intelligence beyond the edge: Inference on intermittent embedded systems,” in Proceedings of the 24th ASPLOS.   ACM, 2019, p. 199–213.
  19. C. Alippi, G. Anastasi, M. Di Francesco, and M. Roveri, “Energy management in wireless sensor networks with energy-hungry sensors,” IEEE Instrumentation & Measurement Magazine, vol. 12, no. 2, pp. 16–23, 2009.
  20. B. Kathirgamanathan and P. Cunningham, “A feature selection method for multi-dimension time-series data,” in Advanced Analytics and Learning on Temporal Data: 5th ECML PKDD Workshop, AALTD 2020, Revised Selected Papers 6.   Springer, 2020, pp. 220–231.
  21. B. Dhariyal, T. L. Nguyen, and G. Ifrim, “Fast Channel Selection for Scalable Multivariate Time Series Classification,” in International Workshop on Advanced Analytics and Learning on Temporal Data.   Springer, 2021, pp. 36–54.
  22. A. P. Ruiz, M. Flynn, J. Large, M. Middlehurst, and A. Bagnall, “The Great Multivariate Time Series Classification Bake off: A Review and Experimental Evaluation of Recent Algorithmic Advances,” Data Mining and Knowledge Discovery, vol. 35, no. 2, pp. 401–449, 2021.
  23. H. Yoon, K. Yang, and C. Shahabi, “Feature subset selection and feature ranking for multivariate time series,” IEEE Transactions on Knowledge and Data Engineering, vol. 17, no. 9, pp. 1186–1198, 2005.
  24. C. Alippi, G. Anastasi, M. Di Francesco, and M. Roveri, “An Adaptive Sampling Algorithm for Effective Energy Management in Wireless Sensor Networks With Energy-Hungry Sensors,” IEEE Transactions on Instrumentation and Measurement, vol. 59, no. 2, pp. 335–344, 2010.
  25. R. Willett, A. Martin, and R. Nowak, “Backcasting: Adaptive Sampling for Sensor Networks,” in Proceedings of the 3rd International Symposium on Information Processing in Sensor Networks, 2004, p. 124–133.
  26. B. Gedik, L. Liu, and P. S. Yu, “ASAP: An Adaptive Sampling Approach to Data Collection in Sensor Networks,” IEEE Transactions on Parallel and Distributed Systems, vol. 18, no. 12, pp. 1766–1783, 2007.
  27. M. F. Ghalwash and Z. Obradovic, “Early classification of multivariate temporal observations by extraction of interpretable shapelets,” BMC bioinformatics, vol. 13, pp. 1–12, 2012.
  28. G. He, Y. Duan, R. Peng, X. Jing, T. Qian, and L. Wang, “Early classification on multivariate time series,” Neurocomputing, vol. 149, pp. 777–787, 2015.
  29. K. Li, S. Li, and Y. Fu, “Early classification of ongoing observation,” in 2014 IEEE ICDM, 2014, pp. 310–319.
  30. A. Gupta, H. P. Gupta, B. Biswas, and T. Dutta, “A fault-tolerant early classification approach for human activities using multivariate time series,” IEEE Transactions on Mobile Computing, vol. 20, no. 5, pp. 1747–1760, 2020.
  31. A. Gupta, H. P. Gupta, B. Biswas, and T. Dutta, “A divide-and-conquer–based early classification approach for multivariate time series with different sampling rate components in IoT,” ACM Transactions on Internet of Things, vol. 1, no. 2, pp. 1–21, 2020.
  32. H. Huang, C. Liu, and V. S. Tseng, “Multivariate time series early classification using multi-domain deep neural network,” in 5th IEEE DSAA, 2018, pp. 90–98.
  33. E.-Y. Hsu, C.-L. Liu, and V. S. Tseng, “Multivariate time series early classification with interpretability using deep learning and attention mechanism,” in PAKDD, Proceedings, Part III 23.   Springer, 2019, pp. 541–553.
  34. C. Martinez, G. Perrin, E. Ramasso, and M. Rombaut, “A deep reinforcement learning approach for early classification of time series,” in 26th EUSIPCO.   IEEE, 2018, pp. 2030–2034.
  35. C. Martinez, E. Ramasso, G. Perrin, and M. Rombaut, “Adaptive early classification of temporal sequences using deep reinforcement learning,” Knowledge-Based Systems, vol. 190, p. 105290, 2020.
  36. V. Mnih, K. Kavukcuoglu, D. Silver, A. A. Rusu, J. Veness, M. G. Bellemare, A. Graves, M. Riedmiller, A. K. Fidjeland, G. Ostrovski et al., “Human-level control through deep reinforcement learning,” nature, vol. 518, no. 7540, pp. 529–533, 2015.
  37. T. Hartvigsen, C. Sen, X. Kong, and E. Rundensteiner, “Recurrent halting chain for early multi-label classification,” in Proceedings of the 26th ACM SIGKDD KDD, 2020, pp. 1382–1392.
  38. T. Hartvigsen, W. Gerych, J. Thadajarassiri, X. Kong, and E. Rundensteiner, “Stop&Hop: Early Classification of Irregular Time Series,” in Proceedings of the 31st ACM CIKM, 2022, pp. 696–705.
  39. R. J. Williams, “Simple statistical gradient-following algorithms for connectionist reinforcement learning,” Reinforcement learning, pp. 5–32, 1992.
  40. Y. Wu, Z. Wang, Z. Jia, Y. Shi, and J. Hu, “Intermittent inference with nonuniformly compressed multi-exit neural network for energy harvesting powered devices,” in 2020 57th ACM/IEEE Design Automation Conference (DAC).   IEEE, 2020, pp. 1–6.
  41. U. Mori, A. Mendiburu, S. Dasgupta, and J. A. Lozano, “Early classification of time series by simultaneously optimizing the accuracy and earliness,” IEEE Transactions on Neural Networks and Learning Systems, vol. 29, no. 10, pp. 4569–4578, 2017.
  42. A. Dempster, F. Petitjean, and G. I. Webb, “ROCKET: exceptionally fast and accurate time series classification using random convolutional kernels,” Data Mining and Knowledge Discovery, vol. 34, no. 5, pp. 1454–1495, 2020.
  43. C. W. Tan, A. Dempster, C. Bergmeir, and G. I. Webb, “MultiRocket: multiple pooling operators and transformations for fast and effective time series classification,” Data Mining and Knowledge Discovery, vol. 36, no. 5, pp. 1623–1646, 2022.
  44. L. Pantiskas, K. Verstoep, M. Hoogendoorn, and H. Bal, “Taking ROCKET on an Efficiency Mission: Multivariate Time Series Classification with LightWaveS,” in 18th IEEE DCOSS, 2022, pp. 149–152.
  45. P.-W. Chou, D. Maturana, and S. Scherer, “Improving stochastic policy gradients in continuous control with deep reinforcement learning using the beta distribution,” in Proceedings of the 34th ICML, vol. 70, 2017, pp. 834–843.
  46. H. Ismail Fawaz, B. Lucas, G. Forestier, C. Pelletier, D. F. Schmidt, J. Weber, G. I. Webb, L. Idoumghar, P.-A. Muller, and F. Petitjean, “Inceptiontime: Finding alexnet for time series classification,” Data Mining and Knowledge Discovery, vol. 34, no. 6, pp. 1936–1962, 2020.
  47. Z. Wang, W. Yan, and T. Oates, “Time series classification from scratch with deep neural networks: A strong baseline,” in 2017 IJCNN.   IEEE, 2017, pp. 1578–1585.
  48. N. Srivastava, G. E. Hinton, A. Krizhevsky, I. Sutskever, and R. Salakhutdinov, “Dropout: A Simple Way to Prevent Neural Networks from Overfitting,” Journal of Machine Learning Research, vol. 15, no. 1, pp. 1929–1958, 2014.
  49. V. Petsiuk, A. Das, and K. Saenko, “Rise: Randomized input sampling for explanation of black-box models,” arXiv preprint arXiv:1806.07421, 2018.
  50. J. Crabbé and M. Van Der Schaar, “Explaining Time Series Predictions with Dynamic Masks,” in Proceedings of the 38th ICML, vol. 139, 2021, pp. 2166–2177.
  51. D. Mercier, A. Dengel, and S. Ahmed, “TimeREISE: Time Series Randomized Evolving Input Sample Explanation,” Sensors, vol. 22, no. 11, 2022.
  52. J. Schulman, P. Moritz, S. Levine, M. Jordan, and P. Abbeel, “High-dimensional continuous control using generalized advantage estimation,” arXiv preprint arXiv:1506.02438, 2015.
  53. H. A. Dau, A. Bagnall, K. Kamgar, C.-C. M. Yeh, Y. Zhu, S. Gharghabi, C. A. Ratanamahatana, and E. Keogh, “The ucr time series archive,” IEEE/CAA Journal of Automatica Sinica, vol. 6, no. 6, pp. 1293–1305, 2019.
  54. A. Bagnall, H. A. Dau, J. Lines, M. Flynn, J. Large, A. Bostrom, P. Southam, and E. Keogh, “The UEA multivariate time series classification archive, 2018,” arXiv:1811.00075 [cs, stat], 2018.
  55. “Machinery Fault Database [Online],” http://www02.smt.ufrj.br/~offshore/mfs/page_01.html.
  56. “Case Western Reserve University Bearing Data Center,” https://engineering.case.edu/bearingdatacenter.
  57. H. Bal, D. Epema, C. de Laat, R. van Nieuwpoort, J. Romein, F. Seinstra, C. Snoek, and H. Wijshoff, “A Medium-Scale Distributed System for Computer Science Research: Infrastructure for the Long Term,” Computer, vol. 49, no. 5, pp. 54–63, 2016.
  58. A. Paszke, S. Gross, F. Massa, A. Lerer, J. Bradbury, G. Chanan, T. Killeen, Z. Lin, N. Gimelshein, L. Antiga et al., “Pytorch: An imperative style, high-performance deep learning library,” Advances in Neural Information Processing Systems, vol. 32, 2019.
  59. I. Oguiza, “tsai - a state-of-the-art deep learning library for time series and sequential data,” Github, 2022. [Online]. Available: https://github.com/timeseriesAI/tsai
  60. A. Sawada, T. Miyagawa, A. Ebihara, S. Yachida, and T. Hosoi, “Convolutional neural networks for time-dependent classification of variable-length time series,” in IJCNN.   IEEE, 2022, pp. 1–8.
  61. R. Islam, P. Henderson, M. Gomrokchi, and D. Precup, “Reproducibility of benchmarked deep reinforcement learning tasks for continuous control,” arXiv preprint arXiv:1708.04133, 2017.
  62. Y. Duan, X. Chen, R. Houthooft, J. Schulman, and P. Abbeel, “Benchmarking deep reinforcement learning for continuous control,” in ICML.   PMLR, 2016, pp. 1329–1338.
  63. G. Gao, Q. Gao, X. Yang, M. Pajic, and M. Chi, “A reinforcement learning-informed pattern mining framework for multivariate time series classification,” in Proceedings of the 31st IJCAI, 2022, pp. 2994–3000.
Citations (2)
View on Semantic Scholar

Summary

We haven't generated a summary for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Now