Abstract
Long short-term memory (LSTM) neural network has been widely studied and applied in the real world. To obtain the LSTM neural network with better accuracy and more appropriate structure, the hybrid coding particle swarm optimization (HCPSO) algorithm is proposed. Firstly, the hybrid coding scheme is developed to represent the weights and structure of LSTM neural network, simultaneously. Then, the novel update mechanism is proposed to adjust the position of particles. Meanwhile, the discrete update strategy (DUS) and adaptive nonlinear moderate random search strategy (ANMRS) are proposed to enhance the convergence and global search capability of HCPSO, respectively. Finally, the effectiveness of HCPSO is demonstrated by multiple numerical examples. The experiment results show that the proposed HCPSO algorithm is more competitive in optimizing LSTM neural networks than other algorithms.
Similar content being viewed by others
References
Greff K, Srivastava RK, Koutnik J, Steunebrink BR, Schmidhuber J (2017) LSTM: a search space odyssey. IEEE Transac Neural Netw Learn Syst 28(10):2222–2232. https://doi.org/10.1109/tnnls.2016.2582924
Wen S, Wang Y, Tang Y, Xu Y, Li P, Zhao T (2019) Real-time identification of power fluctuations based on LSTM recurrent neural network: a case study on singapore power system. IEEE Trans Industr Inf 15(9):5266–5275. https://doi.org/10.1109/TII.2019.2910416
Dong Y, Zhang H, Wang C, Zhou X (2021a) Wind power forecasting based on stacking ensemble model, decomposition and intelligent optimization algorithm. Neurocomputing 462:169–184. https://doi.org/10.1016/j.neucom.2021.07.084
Dong Y, Zhang H, Wang C, Zhou X (2021b) A novel hybrid model based on Bernstein polynomial with mixture of gaussians for wind power forecasting. Appl Energy . https://doi.org/10.1016/j.apenergy.2021.116545
Zhao K, Li L, Cai S, Sun R (2019) An optimized time series combined forecasting method based on neural networks. J Supercomput 76(4):2986–3012. https://doi.org/10.1007/s11227-019-03064-5
Paoletti ME, Haut JM, Plaza J, Plaza A (2020) Scalable recurrent neural network for hyperspectral image classification. J Supercomput 76(11):8866–8882. https://doi.org/10.1007/s11227-020-03187-0
Hochreiter S, Schmidhuber J (1997) Long short-term memory. Neural Comput 9(8):1735–80. https://doi.org/10.1162/neco.1997.9.8.1735
Bengio Y, Simard P, Frasconi P (1994) Learning long-term dependencies with gradient descent is difficult. IEEE Trans Neural Netw 5(2):157–166. https://doi.org/10.1109/72.279181
Mazumdar J, Harley RG (2008) Recurrent neural networks trained with backpropagation through time algorithm to estimate nonlinear load harmonic currents. IEEE Trans Industr Electron 55(9):3484–3491. https://doi.org/10.1109/tie.2008.925315
Puskorius GV, Feldkamp LA (1994) Neurocontrol of nonlinear dynamical-systems with kalman filter trained recurrent networks. IEEE Trans Neural Netw 5(2):279–297. https://doi.org/10.1109/72.279191
Ergen T, Kozat SS (2018) Efficient online learning algorithms based on LSTM neural networks. IEEE Trans Neural Netw Learn Syst 29(8):3772–3783. https://doi.org/10.1109/tnnls.2017.2741598
Wang ZSLJMLJWZFMQ, Wang, (2019) E-LSTM: An Efficient Hardware Architecture for Long Short-Term. IEEE J Emerging Selected Topics Circuits Syst. https://doi.org/10.1109/jetcas.2019.2911739
Sagheer A, Kotb M (2019) Time series forecasting of petroleum production using deep LSTM recurrent networks. Neurocomputing 323:203–213. https://doi.org/10.1016/j.neucom.2018.09.082
ElSaid A, El Jamiy F, Higgins J, Wild B, Desell T (2018) Optimizing long short-term memory recurrent neural networks using ant colony optimization to predict turbine engine vibration. Appl Soft Comput 73:969–991. https://doi.org/10.1016/j.asoc.2018.09.013
Gundu V, Simon SP (2021) PSO-LSTM for short term forecast of heterogeneous time series electricity price signals. J Ambient Intell Humaniz Comput 12(2):2375–2385. https://doi.org/10.1007/s12652-020-02353-9
Yu J, Wang S, Xi L (2008) Evolving artificial neural networks using an improved PSO and DPSO. Neurocomputing 71(4–6):1054–1060. https://doi.org/10.1016/j.neucom.2007.10.013
Zhao L, Qian F (2011) Tuning the structure and parameters of a neural network using cooperative binary-real particle swarm optimization. Expert Syst Appl 38(5):4972–4977. https://doi.org/10.1016/j.eswa.2010.09.154
Kiranyaz S, Ince T, Yildirim A, Gabbouj M (2009) Evolutionary artificial neural networks by multi-dimensional particle swarm optimization. Neural Netw 22(10):1448–62. https://doi.org/10.1016/j.neunet.2009.05.013
Han HG, Lu W, Hou Y, Qiao JF (2018) An adaptive-PSO-based self-organizing rbf neural network. IEEE TransNeural Netw Learn Syst 29(1):104–117. https://doi.org/10.1109/tnnls.2016.2616413
Gao H, Xu WB (2011) A new particle swarm algorithm and its globally convergent modifications. IEEE Trans Syst Man and Cybernet Part B-Cybernet 41(5):1334–1351. https://doi.org/10.1109/tsmcb.2011.2144582
Yaghini M, Khoshraftar MM, Fallahi M (2013) A hybrid algorithm for artificial neural network training. Eng Appl Artif Intell 26(1):293–301. https://doi.org/10.1016/j.engappai.2012.01.023
Aljarah I, Faris H, Mirjalili S (2016) Optimizing connection weights in neural networks using the whale optimization algorithm. Soft Comput 22(1):1–15. https://doi.org/10.1007/s00500-016-2442-1
Mirjalili S, Mohd Hashim SZ, Moradian Sardroudi H (2012) Training feedforward neural networks using hybrid particle swarm optimization and gravitational search algorithm. Appl Math Comput 218(22):11125–11137. https://doi.org/10.1016/j.amc.2012.04.069
Shao BL, Li ML, Zhao Y, Bian GQ (2019) Nickel Price Forecast Based on the LSTM Neural Network Optimized by the Improved PSO algorithm. Math Probl Eng 2019:15. https://doi.org/10.1155/2019/1934796
Sheikhan M, Abbasnezhad Arabi M, Gharavian D (2015) Structure and weights optimisation of a modified elman network emotion classifier using hybrid computational intelligence algorithms: a comparative study. Connect Sci 27(4):340–357. https://doi.org/10.1080/09540091.2015.1080224
Zhang L, Li H, Kong XG (2019) Evolving feedforward artificial neural networks using a two-stage approach. Neurocomputing 360:25–36. https://doi.org/10.1016/j.neucom.2019.03.097
Kennedy J, Eberhart R, IEEE, IEEE, IEEE, IEEE (1995) Particle swarm optimization. 1995 IEEE Int Conf Neural Netw Proc. https://doi.org/10.1109/icnn.1995.488968
Rai P, Hanmandlu M (2016) Gaussian Function-Based Particle Swarm Optimization. Springer Singapore, Proceedings of Fifth International Conference on Soft Computing for Problem Solving, pp 57–67
Yang C, Qiao J, Wang L, Zhu X (2018) Dynamical regularized echo state network for time series prediction. Neural Comput Appl 31(10):6781–6794. https://doi.org/10.1007/s00521-018-3488-z
Qiao J, Quan L, Yang C (2020) Design of modeling error PDF based fuzzy neural network for effluent ammonia nitrogen prediction. Appl Soft Comput . https://doi.org/10.1016/j.asoc.2020.106239
Kennedy J, Eberhart RC, IEEE, (1997) A discrete binary version of the particle swarm algorithm, pp 4104–4108. IEEE Int Conf Syst Man, and Cybernatics, Conf Proc. https://doi.org/10.1109/icsmc.1997.637339
Yang C, Qiao J, Ahmad Z, Nie K, Wang L (2019) Online sequential echo state network with sparse RLS algorithm for time series prediction. Neural Netw 118:32–42. https://doi.org/10.1016/j.neunet.2019.05.006
Acknowledgements
This work was supported by National Natural Science Foundation of China (62021003, 61890930-5 and 61973010), National Natural Science Foundation of Beijing (4202006) and Major Science and Technology Program for Water Pollution Control and Treatment of China (2018ZX07111-005).
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Appendices
Appendix A: The description of ODUS and MRS
For ODUS of HCPSO-LSTM-ODUS, the update of the particle position is as follows:
with
For MRS of HCPSO-LSTM-MRS, the update of the particle position is as follows:
where \(\beta _{\mathrm{max}}\) and \(\beta _{\mathrm{min}}\) are the upper and lower limits of \(\beta \), respectively, and they take values in the range of \([0.1,\,1.0]\).
Appendix B: Nomenclature and abbreviations
Rights and permissions
About this article
Cite this article
Chen, Z., Yang, C. & Qiao, J. The optimal design and application of LSTM neural network based on the hybrid coding PSO algorithm. J Supercomput 78, 7227–7259 (2022). https://doi.org/10.1007/s11227-021-04142-3
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11227-021-04142-3