LSTM based Stacked Autoencoder Approach for Time Series Forecasting

K.N. Singh; Kamal Sharma; G. Avinash; Rajeev Ranjan Kumar; Mrinmoy Ray; Ramasubramanian V.; Achal Lama; S.B. Lal

doi:10.56093/jisas.v77i01.171563

Authors

K.N. Singh ICAR-Indian Agricultural Statistics Research Institute, New Delhi
Kamal Sharma ICAR-Indian Agricultural Statistics Research Institute, New Delhi
G. Avinash ICAR-Indian Agricultural Statistics Research Institute, New Delhi
Rajeev Ranjan Kumar ICAR-Indian Agricultural Statistics Research Institute, New Delhi
Mrinmoy Ray ICAR-Indian Agricultural Statistics Research Institute, New Delhi
Ramasubramanian V. ICAR-Indian Agricultural Statistics Research Institute, New Delhi
Achal Lama ICAR-Indian Agricultural Statistics Research Institute, New Delhi
S.B. Lal ICAR-Indian Agricultural Statistics Research Institute, New Delhi

https://doi.org/10.56093/jisas.v77i01.171563

Keywords:

Deep learning, LSTM, Autoencoder, Gated Recurrent Unit (GRU), Stock price.

Abstract

This study proposes a novel approach for multi-step time series forecasting using a stacked long-short term memory (LSTM) sequence-to-sequence autoencoder (LSTM-SAE) to handle the volatility of edible oil prices in the Indian market. The approach was implemented on Ruchi Soya Ltd. stock price dataset and compared with other deep learning models like Gated Recurrent Unit (GRU), LSTM, and Bi-directional LSTM. The LSTM SAE outperformed other models in closing price prediction based on evaluation metrics like Root Mean Square Error (RMSE) and Mean Absolute Percentage Error (MAPE). The proposed approach has significant implications for stakeholders in the edible oil and oilseeds industry, including farmers, traders, and policymakers

Downloads

Download data is not yet available.

References

Araujo, R.D.A., Oliveira, A.L. and Meira, S. (2015). A hybrid model for high-frequency stock market forecasting. Expert Systems with Applications, 42(8), 4081-4096.

Bao, W., Yue, J. and Rao, Y. (2017). A deep learning framework for financial time series using stacked autoencoders and long-short term memory. PloS one, 12(7), e0180944.

Cho, K., Van Merrienboer, B., Bahdanau, D. and Bengio, Y. (2014). On the properties of neural machine translation: Encoder-decoder approaches. arXiv preprint arXiv:1409.1259.

Chong, E., Han, C. and Park, F.C. (2017). Deep learning networks for stock market analysis and prediction: Methodology, data representations, and case studies. Expert Systems with Applications, 83, 187-205.

Gianniotis, N., Kugler, S.D., Tino, P. and Polsterer, K.L. (2016). Model-coupled Autoencoder for time series visualisation. Neurocomputing, 192, 139-146.

Guo, Y., Liao, W., Wang, Q., Yu, L., Ji, T. and Li, P. (2018, November). Multidimensional time series anomaly detection: A gru-based gaussian mixture variational autoencoder approach. In Asian Conference on Machine Learning (97-112).

Haykin, S. (2009). Neural networks and learning machines, 3/E. Pearson Education India.

Hochreiter, S. and Schmidhuber, J. (1997). Long short-term memory. Neural computation, 9(8), 1735-1780.

Homayouni, H., Ghosh, S., Ray, I., Gondalia, S., Duggan, J. and Kahn, M.G. (2020). An autocorrelation-based lstm-autoencoder for anomaly detection on time-series data. In 2020 IEEE International Conference on Big Data (Big Data) (5068-5077).

Jaiswal, R., Jha, G.K., Kumar, R.R. and Choudhary, K. (2021). Deep long short-term memory-based model for agricultural price forecasting. Neural Computing and Applications, 34, 4661–4676. https://doi.org/10.1007/s00521-021-06621-3

Jiang, S., and Durlofsky, L.J. (2021). Data-space inversion using a recurrent autoencoder for time-series parameterisation. Computational Geosciences, 25, 411-432.

Jung, G., and Choi, S.Y. (2021). Forecasting foreign exchange volatility using deep learning autoencoder-LSTM techniques. Complexity. https://doi.org/10.1155/2021/6647534

Kieu, T., Yang, B., Guo, C., and Jensen, C.S. (2019, August). Outlier Detection for Time Series with Recurrent Autoencoder Ensembles. In IJCAI (2725-2732).

Li, L., Yan, J., Wang, H., and Jin, Y. (2020). Anomaly detection of time series with smoothness-inducing sequential variational autoencoder. IEEE transactions on neural networks and learning systems, 32(3), 1177-1191.

Liguori, A., Markovic, R., Dam, T.T.H., Frisch, J., van Treeck, C., and Causone, F. (2021). Indoor environment data time-series reconstruction using autoencoder neural networks. Building and Environment, 191, 107623.

Liu, P., Sun, X., Han, Y., He, Z., Zhang, W., and Wu, C. (2022). Arrhythmia classification of LSTM autoencoder based on time series anomaly detection. Biomedical Signal Processing and Control, 71, 103228.

Mohanty, D.K., Parida, A.K., and Khuntia, S.S. (2021). Financial market prediction under deep learning framework using autoencoder and kernel extreme learning machine. Applied Soft Computing, 99, 106898.

Nguyen, N., and Quanz, B. (2021, May). Temporal latent auto-encoder: A method for probabilistic multivariate time series forecasting. In Proceedings of the AAAI Conference on Artificial Intelligence, 35(10), 9117-9125.

Noering, F.K.D., Schroeder, Y., Jonas, K. and Klawonn, F. (2021). Pattern discovery in time series using Autoencoder in comparison to nonlearning approaches. Integrated Computer-Aided Engineering, 28(3), 237-256.

Pan, Z., Wang, Y., Wang, K., Chen, H., Yang, C. and Gui, W. (2022). Imputation of missing values in time series using an adaptive learned median filled deep autoencoder. IEEE Transactions on Cybernetics, 53(2), 695-706. DOI: 10.1109/TCYB.2022.3167995.

Provotar, O.I., Linder, Y.M. and Veres, M.M. (2019, December). Unsupervised anomaly detection in time series using lstm-based autoencoders. International Conference on Advanced Trends in Information Theory (ATIT), 513-517.

Richard, G., Grossin, B., Germaine, G., Hebrail, G. and de Moliner, A. (2020). Autoencoder-based time series clustering with energy applications. arXiv preprint arXiv:2002.03624.

Sagheer, A., and Kotb, M. (2019). Unsupervised pre-training of a deep LSTM-based stacked autoencoder for multivariate time series forecasting problems. Scientific reports, 9(1), 1-16.

Shumway, R.H., Stoffer, D.S. and Stoffer, D.S. (2000). Time series analysis and its applications (3). New York: Springer.

Terefe, T., Devanne, M., Weber, J., Hailemariam, D. and Forestier, G. (2020, November). Time series averaging using multi-tasking autoencoder. 32nd International Conference on Tools with Artificial Intelligence (ICTAI), 1065-1072.

Von Schleinitz, J., Graf, M., Trutschnig, W. and Schröder, A. (2021). VASP: An autoencoder-based approach for multivariate anomaly detection and robust time series prediction with application in motorsport. Engineering Applications of Artificial Intelligence, 104, 104354.

Zhang, C., Li, S., Zhang, H., and Chen, Y. (2019). Velc: A new variational autoencoder based model for time series anomaly detection. arXiv preprint arXiv:1907.01702.

Zhang, J., and Dai, Q. (2022). Latent adversarial regularised Autoencoder for high-dimensional probabilistic time series prediction. Neural Networks, 155, 383-397.

Zhao, X., Han, X., Su, W., and Yan, Z. (2019, November). Time series prediction method based on convolutional Autoencoder and LSTM. In 2019 Chinese Automation Congress (CAC), 5790-5793.

Zhou, H., Yu, K., Zhang, X., Wu, G., and Yazidi, A. (2022). Contrastive Autoencoder for anomaly detection in multivariate time series. Information Sciences, 610, 266-280.

Wei, Y., Jang-Jaccard, J., Xu, W., Sabrina, F., Camtepe, S. and Boulic, M. (2023). Lstm-autoencoder based anomaly detection for indoor air quality time series data. Sensors. DOI: https://doi.org/10.1109/JSEN.2022.3230361

Xie, L., and Yu, S. (2021). Unsupervised feature extraction with convolutional Autoencoder with application to daily stock market prediction. Concurrency and Computation: Practice and Experience, 33(16), 6282.

Wu, K., Liu, J., Liu, P., and Yang, S. (2019). Time series prediction using sparse Autoencoder and high-order fuzzy cognitive maps. IEEE transactions on fuzzy systems, 28(12), 3110-3121.

Xu, X., and Ren, W. (2022). A hybrid model of stacked Autoencoder and modified particle swarm optimisation for multivariate chaotic time series forecasting. Applied Soft Computing, 116, 108321.

Yin, C., Zhang, S., Wang, J., and Xiong, N.N. (2020). Anomaly detection based on convolutional recurrent Autoencoder for IoT time series. IEEE Transactions on Systems, Man, and Cybernetics: Systems, 52(1), 112-122.

Yu, W., Kim, I.Y., and Mechefske, C. (2021). Analysis of different RNN autoencoder variants for time series classification and machine prognostics. Mechanical Systems and Signal Processing, 149, 107322.

Zha, M. (2022). Time series generation with masked Autoencoder. arXiv preprint arXiv:2201.07006.