Improving demand forecasting with LSTM by taking into account the seasonality of data

Document Type: Research Paper

Authors

Faculty of Information Technology and Computer Engineering, Azarbaijan Shahid Madani University, Tabriz, Iran.

Abstract

Demand forecasting is a vital task for firms to manage the optimum quantity of raw material and products. The demand forecasting task can be formulated as a time series forecasting problem by measuring historical demand data at equal intervals. Demand time series usually exhibit a seasonal pattern. The principle idea of this study is to propose a method that predicts the demand for every different season using a specialized forecaster. In this study, we test our proposal using the Long Short-Term Memory (LSTM) which is a deep learning technique for time series forecasting. Specifically, the proposed method instead of learning an LSTM model using the whole demand data builds a specialized LSTM model corresponding to each season. The proposed method is evaluated using different topologies of the LSTM model. The results of experiments indicated that the proposed method outperforms the regular method considering the performance measures. The proposed method can be used in other domains for demand forecasting.

Keywords

Main Subjects


[1]     Kumar, A., Shankar, R., & Aljohani, N. R. (2020). A big data driven framework for demand-driven forecasting with effects of marketing-mix variables. Industrial marketing management90, 493-507.
[2]     Villegas, M. A., Pedregal, D. J., & Trapero, J. R. (2018). A support vector machine for model selection in demand forecasting applications. Computers & industrial engineering, 121, 1-7. https://doi.org/10.1016/j.cie.2018.04.042.
[3]     Johannesen, N. J., Kolhe, M., & Goodwin, M. (2019). Relative evaluation of regression tools for urban area electrical energy demand forecasting. Journal of cleaner production, 218, 555-564.
[4]     Law, R., Li, G., Fong, D. K. C., & Han, X. (2019). Tourism demand forecasting: A deep learning approach. Annals of tourism research, 75, 410-423
[5]     Olaniyi, A. A., Adedotun, K. O., & Samuel, O. A. (2018). Forecasting methods for domestic air passenger demand in Nigeria. Journal of applied research on industrial engineering, 5(2), 146-155. 10.22105/jarie.2018.133561.1038
[6]     Bandara, K., Bergmeir, C., & Smyl, S. (2020). Forecasting across time series databases using recurrent neural networks on groups of similar series: A clustering approach. Expert systems with applications, 140, 112896. https://doi.org/10.1016/j.eswa.2019.112896
[7]     Sagheer, A., & Kotb, M. (2019). Time series forecasting of petroleum production using deep LSTM recurrent networks. Neurocomputing, 323, 203-213.
[8]     Panigrahi, S., & Behera, H. S. (2017). A hybrid ETS–ANN model for time series forecasting. Engineering applications of artificial intelligence, 66, 49-59.
[9]     Parmezan, A. R. S., Souza, V. M., & Batista, G. E. (2019). Evaluation of statistical and machine learning models for time series prediction: Identifying the state-of-the-art and the best conditions for the use of each model. Information sciences, 484, 302-337.
[10] Hochreiter, S., & Schmidhuber, J. (1997). Long short-term memory. Neural computation, 9(8), 1735-1780.
[11] Wu, Y., Yuan, M., Dong, S., Lin, L., & Liu, Y. (2018). Remaining useful life estimation of engineered systems using vanilla LSTM neural networks. Neurocomputing, 275, 167-179. https://doi.org/10.1016/j.neucom.2017.05.063.
[12] Khashei, M., & Bijari, M. (2011). A novel hybridization of artificial neural networks and ARIMA models for time series forecasting. Applied soft computing, 11(2), 2664-2675.
[13] Murray, P. W., Agard, B., & Barajas, M. A. (2018). Forecast of individual customer’s demand from a large and noisy dataset. Computers & industrial engineering, 118, 33-43.
[14] Abbasimehr, H., & Shabani, M. (2020). A new framework for predicting customer behavior in terms of RFM by considering the temporal aspect based on time series techniques. Journal of ambient intelligence and humanized computing. 10.1007/s12652-020-02015-w.
[15] Fischer, T., & Krauss, C. (2018). Deep learning with long short-term memory networks for financial market predictions. European journal of operational research, 270(2), 654-669.
[16] Bengio, Y., Simard, P., & Frasconi, P. (1994). Learning long-term dependencies with gradient descent is difficult. IEEE transactions on neural networks, 5(2), 157-166.
[17] Xin, W. A. N. G., Ji, W. U., Chao, L. I. U., Haiyan, Y. A. N. G., Yanli, D. U., & Wensheng, N. I. U. (2018). Exploring LSTM based recurrent neural network for failure time series prediction. Journal of beijing university of aeronautics and astronautics, 44(4), 772-784.
[18] Abbasimehr, H., Shabani, M., & Yousefi, M. (2020). An optimized model using LSTM network for demand forecasting. Computers & industrial engineering. https://doi.org/10.1016/j.cie.2020.106435.
[19] Shankar, S., Ilavarasan, P. V., Punia, S., & Singh, S. P. (2019). Forecasting container throughput with long short-term memory networks. Industrial management & data systems. 120(3), 425-441. 10.1108/IMDS-07-2019-0370.
[20] Ke, J., Zheng, H., Yang, H., & Chen, X. M. (2017). Short-term forecasting of passenger demand under on-demand ride services: A spatio-temporal deep learning approach. Transportation research part C: Emerging technologies, 85, 591-608.
[21] Pan, B., Yuan, D., Sun, W., Liang, C., & Li, D. (2018, June). A novel LSTM-Based Daily Airline Demand Forecasting Method Using Vertical and Horizontal Time series. Pacific-Asia conference on knowledge discovery and data mining (pp. 168-173). Springer, Cham.
[22] Bedi, J., & Toshniwal, D. (2018). Empirical mode decomposition based deep learning for electricity demand forecasting. IEEE access, 6, 49144-49156.
[23] Bouktif, S., Fiaz, A., Ouni, A., & Serhani, M. A. (2018). Optimal deep learning lstm model for electric load forecasting using feature selection and genetic algorithm: Comparison with machine learning approaches. Energies11(7), 1636.
[24] Bedi, J., & Toshniwal, D. (2019). Deep learning framework to forecast electricity demand. Applied energy, 238, 1312-1326.
[25] Su, H., Zio, E., Zhang, J., Xu, M., Li, X., & Zhang, Z. (2019). A hybrid hourly natural gas demand forecasting method based on the integration of wavelet transform and enhanced Deep-RNN model. Energy. 178, 585-597. https://doi.org/10.1016/j.energy.2019.04.167
[26] Tan, M., Yuan, S., Li, S., Su, Y., Li, H., & He, F. (2019). Ultra-short-term industrial power demand forecasting using LSTM based hybrid ensemble learning. IEEE transactions on power systems, 35(4), 2937-2948. 10.1109/TPWRS.2019.2963109
[27] Kulshrestha, A., Krishnaswamy, V., & Sharma, M. (2020). Bayesian BILSTM approach for tourism demand forecasting. Annals of tourism research, 83, 102925. https://doi.org/10.1016/j.annals.2020.102925
[28] Punia, S., Nikolopoulos, K., Singh, S. P., Madaan, J. K., & Litsiou, K. (2020). Deep learning with long short-term memory networks and random forests for demand forecasting in multi-channel retail. International journal of production research, 1-16. 10.1080/00207543.2020.1735666
[29] Bui, V., Kim, J., & Jang, Y. M. (2020, February). Power demand forecasting using long short-term memory neural network based smart grid. 2020 international conference on artificial intelligence in information and communication (ICAIIC) (pp. 388-391). IEEE.
[30] Wu, D. C. W., Ji, L., He, K., & Tso, K. F. G. (2020). Forecasting tourist daily arrivals with a hybrid Sarima–Lstm approach. Journal of hospitality & tourism research. https://doi.org/10.1177/1096348020934046
[31] Chollet, F. (2015). Keras.  Retrived January 12, 2020 from https://github.com/fchollet/keras.
[32] Graves, A. (2013). Generating sequences with recurrent neural networks. https://arxiv.org/
[33] Jiawei Han, M. K., & Pei, J. (2011). Data mining: concepts and techniques: concepts and techniques. Waltham, USA: Elsevier Science.
[34] Martínez, F., Frías, M. P., Pérez-Godoy, M. D., & Rivera, A. J. (2018). Dealing with seasonality by narrowing the training set in time series forecasting with kNN. Expert systems with applications103, 38-48.