Adaptive Scaling for U-Net in Time Series Classification
Pages 309 - 320
Abstract
Convolutional Neural Networks such as U-Net are recently getting popular among researchers in many applications, such as Biomedical Image Segmentation. U-Net is one of the popular deep Convolutional Neural Networks which first contracts the input image using pooling layers and then upscales the feature maps before classifying them. In this paper, we explore the performance of adaptive scaling for U-Net in time series classification. Also, to improve performance, we extract features from the trained U-Net model and use ensemble deep Random Vector Functional Link (edRVFL) to classify them. Experiments on 55 large UCR datasets reveal that adaptive scaling improves the performance of U-Net in time series classification. Also, using edRVFL on extracted features from the trained U-Net model enhances performance. Consequently, our U-Net-edRVFL classifier outperforms other time series classification methods.
References
[1]
Bagnall A, Lines J, Hills J, and Bostrom A Time-series classification with cote: the collective of transformation-based ensembles IEEE Trans. Knowl. Data Eng. 2015 27 9 2522-2535
[2]
Bostrom A and Bagnall A Hameurlain A, Küng J, Wagner R, Madria S, and Hara T Binary shapelet transform for multiclass time series classification Transactions on Large-Scale Data- and Knowledge-Centered Systems XXXII 2017 Heidelberg Springer 24-46
[3]
Chen, Y., et al.: The UCR time series classification archive, July 2015. www.cs.ucr.edu/~eamonn/time_series_data/
[4]
Cheng WX, Suganthan P, and Katuwal R Time series classification using diversified ensemble deep random vector functional link and resnet features Appl. Soft Comput. 2021 112
[5]
Dash Y, Mishra SK, Sahany S, and Panigrahi BK Indian summer monsoon rainfall prediction: a comparison of iterative and non-iterative approaches Appl. Soft Comput. 2018 70 1122-1134
[6]
Dempster A, Petitjean F, and Webb GI ROCKET: exceptionally fast and accurate time series classification using random convolutional kernels Data Min. Knowl. Disc. 2020 34 5 1454-1495
[7]
Deng H, Runger G, Tuv E, and Vladimir M A time series forest for classification and feature extraction Inf. Sci. 2013 239 142-153
[8]
Esser, P., Sutter, E., Ommer, B.: A variational U-net for conditional appearance and shape generation. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), June 2018
[9]
Gamboa, J.C.B.: Deep learning for time-series analysis. arXiv preprint arXiv:1701.01887 (2017)
[10]
Ganaie, M., Hu, M., Malik, A., Tanveer, M., Suganthan, P.: Ensemble deep learning: a review. Eng. Appl. Artif. Intell. 115, 105151 (2022).
[11]
Gao, R., Du, L., Suganthan, P.N., Zhou, Q., Yuen, K.F.: Random vector functional link neural network based ensemble deep learning for short-term load forecasting. Expert Syst. Appl. 206, 117784 (2022)., https://www.sciencedirect.com/science/article/pii/S0957417422010545
[12]
Górecki T and Łuczak M Using derivatives in time series classification Data Min. Knowl. Disc. 2013 26 2 310-331
[13]
Hills J, Lines J, Baranauskas E, Mapp J, and Bagnall A Classification of time series by shapelet transformation Data Min. Knowl. Disc. 2014 28 4 851-881
[14]
Hu, B., Rakthanmanon, T., Hao, Y., Evans, S., Lonardi, S., Keogh, E.: Discovering the intrinsic cardinality and dimensionality of time series using mdl. In: 2011 IEEE 11th International Conference on Data Mining, pp. 1086–1091 (2011).
[15]
Ioffe, S., Szegedy, C.: Batch normalization: accelerating deep network training by reducing internal covariate shift. arXiv preprint arXiv:1502.03167 (2015)
[16]
Katuwal R, Suganthan P, and Zhang L An ensemble of decision trees with random vector functional link networks for multi-class classification Appl. Soft Comput. 2018 70 1146-1153
[17]
Kingma, D.P., Ba, J.: Adam: a method for stochastic optimization. arXiv preprint arXiv:1412.6980 (2014)
[18]
Lavangnananda, K., Sawasdimongkol, P.: Neural network classifier of time series: a case study of symbolic representation preprocessing for control chart patterns. In: 2012 8th International Conference on Natural Computation, pp. 344–349 (2012).
[19]
Lin, M., Chen, Q., Yan, S.: Network in network. arXiv preprint arXiv:1312.4400 (2013)
[20]
Lines J and Bagnall A Time series classification with ensembles of elastic distance measures Data Min. Knowl. Disc. 2015 29 3 565-592
[21]
Liu, Z., Cao, Y., Wang, Y., Wang, W.: Computer vision-based concrete crack detection using U-net fully convolutional networks. Autom. Constr. 104, 129–139 (2019)., https://www.sciencedirect.com/science/article/pii/S0926580519301244
[22]
Ma Q, Zhuang W, Shen L, and Cottrell GW Time series classification with echo memory networks Neural Netw. 2019 117 225-239
[23]
Malik, A.K., Gao, R., Ganaie, M.A., Tanveer, M., Suganthan, P.N.: Random vector functional link network: recent developments, applications, and future directions (2022).
[24]
Nair, V., Hinton, G.E.: Rectified linear units improve restricted Boltzmann machines. In: Proceedings of the 27th International Conference on Machine Learning (ICML-10), pp. 807–814 (2010)
[25]
Oktay, O., et al.: Attention U-net: learning where to look for the pancreas. arXiv preprint arXiv:1804.03999 (2018)
[26]
Pao YH and Takefuji Y Functional-link net computing: theory, system architecture, and functionalities Computer 1992 25 5 76-79
[27]
Rajkomar A et al. Scalable and accurate deep learning with electronic health records NPJ Digit. Med. 2018 1 1 18
[28]
Ronneberger O, Fischer P, and Brox T Navab N, Hornegger J, Wells WM, and Frangi AF U-net: convolutional networks for biomedical image segmentation Medical Image Computing and Computer-Assisted Intervention — MICCAI 2015 2015 Cham Springer 234-241
[29]
Schäfer P The boss is concerned with time series classification in the presence of noise Data Min. Knowl. Disc. 2015 29 6 1505-1530
[30]
Shi, Q., Hu, M., Suganthan, P.N., Katuwal, R.: Weighting and pruning based ensemble deep random vector functional link network for tabular data classification. Pattern Recogn. 132, 108879 (2022)., https://www.sciencedirect.com/science/article/pii/S0031320322003600
[31]
Shi Q, Katuwal R, Suganthan P, and Tanveer M Random vector functional link neural network based ensemble deep learning Pattern Recogn. 2021 117
[32]
Shi, Q., Suganthan, P.N., Del Ser, J.: Jointly optimized ensemble deep random vector functional link network for semi-supervised classification. Eng. Appl. Artif. Intell. 115, 105214 (2022)., https://www.sciencedirect.com/science/article/pii/S0952197622002974
[33]
Suganthan PN and Katuwal R On the origins of randomization-based feedforward neural networks Appl. Soft Comput. 2021 105
[34]
Vuković N, Petrović M, and Miljković Z A comprehensive experimental evaluation of orthogonal polynomial expanded random vector functional link neural networks for regression Appl. Soft Comput. 2018 70 1083-1096
[35]
Wang, J., Chen, Y., Hao, S., Peng, X., Hu, L.: Deep learning for sensor-based activity recognition: a survey. Pattern Recogn. Lett. 119, 3–11 (2019)., Deep Learning for Pattern Recognition
[36]
Wang, Z., Yan, W., Oates, T.: Time series classification from scratch with deep neural networks: a strong baseline. In: 2017 International Joint Conference on Neural Networks (IJCNN), pp. 1578–1585, May 2017.
Recommendations
Denoising temporal convolutional recurrent autoencoders for time series classification
AbstractIn this paper, a denoising temporal convolutional recurrent autoencoder (DTCRAE) is proposed to improve the performance of the temporal convolutional network (TCN) on time series classification (TSC). The DTCRAE consists of a TCN ...
Exploiting multi-channels deep convolutional neural networks for multivariate time series classification
Time series classification is related to many different domains, such as health informatics, finance, and bioinformatics. Due to its broad applications, researchers have developed many algorithms for this kind of tasks, e.g., multivariate time series ...
Time series classification based on statistical features
AbstractThis paper presents a statistical feature approach in fully convolutional time series classification (TSC), which is aimed at improving the accuracy and efficiency of TSC. This method is based on fully convolutional neural networks (FCN), and ...
Comments
Information & Contributors
Information
Published In
Nov 2022
659 pages
ISBN:978-3-031-30104-9
DOI:10.1007/978-3-031-30105-6
© The Author(s), under exclusive license to Springer Nature Switzerland AG 2023.
Publisher
Springer-Verlag
Berlin, Heidelberg
Publication History
Published: 13 April 2023
Author Tags
Qualifiers
- Article
Contributors
Other Metrics
Bibliometrics & Citations
Bibliometrics
Article Metrics
- 0Total Citations
- 0Total Downloads
- Downloads (Last 12 months)0
- Downloads (Last 6 weeks)0
Reflects downloads up to 08 Feb 2025