Abstract
We present a computing scheme as a variant of a recently proposed granular recurrent neural network. Being deduced from a generic system of partial differential equations, this variant is able to capture the spatiotemporal variability of some datasets and problems. The convergence of the computing scheme has been formally discussed. Some preliminary numerical experiments were first performed by using synthetic datasets, inferring some particular partial differential equations. Then two application examples were considered (by using publicly available datasets), namely the prediction of dissolved oxygen in surface water simultaneously at different depths (unlike the current literature) and the prediction of the concentration of particulate matter less than 2.5 \(\upmu\)m in diameter at different sites. The numerical results show the potential of the approach for forecast against well-known techniques such as Long Short-Term Memory networks.
Similar content being viewed by others
References
Bessa RJ, Trindade A, Miranda V (2015) Spatial-temporal solar power forecasting for smart grids. IEEE Trans Ind Inf 11(1):232–241
Tascikaraoglu A (2018) Evaluation of spatio-temporal forecasting methods in various smart city applications. Renew Sustain Energy Rev 82:424–435
Wan H, Guo S, Yin K, Liang X, Lin Y (2019) CTS-LSTM: LSTM-based neural networks for correlated time series prediction, in press
Asadi R, Regan AC (2020) A spatio-temporal decomposition based deep neural network for time series forecasting. Appl Soft Comput 87:105963
Xu W, Wang Q, Chen R (2018) Spatio-temporal prediction of crop disease severity for agricultural emergency management based on recurrent neural networks. Geoinformatica 22:363–381
Wetzel RG (2001) Limnology: lake and river ecosystems, 3rd edn. Academic Press, San Diego, CA
Song C et al (2017) Health burden attributable to ambient PM2.5 in China. Environ Pollut 223:575–586
Hull V, Parrella L, Falcucci M (2008) Modelling dissolved oxygen dynamics in coastal lagoons. Ecol Model 211:468–480
Long BT (2020) Inverse algorithm for Streeter-Phelps equation in water pollution control problem. Math Comput Simul 171:119–126
Boano F, Revelli R, Ridolfi L (2006) Stochastic modelling of DO and BOD components in a stream with random inputs. Adv Water Resour 29:1341–1350
Olyaie E, Abyaneh HZ, Mehr AD (2017) A comparative analysis among computational intelligence techniques for dissolved oxygen prediction in Delaware River. Geosci Front 8(3):517–527
Liu Y, Zhang Q, Song L, Chen Y (2019) Attention-based recurrent neural networks for accurate short-term and long-term dissolved oxygen prediction. Comput Elect Agric 165:104964
Frieder CA, Nam SH, Martz TR, Levin LA (2012) High temporal and spatial variability of dissolved oxygen and pH in a nearshore California kelp forest. Biogeoscience 9:3917–3930
Noges T, Kangro K (2005) Primary production of phytoplankton in a strongly stratified temperate lake. Hydrobiologia 547:105–122
Keshtegar B, Heddam S (2018) Modeling daily dissolved oxygen concentration using modified response surface method and artificial neural network: a comparative study. Neural Comput Appl 30:2995–3006
Antanasijevic D, Pocajt V, Peric-Grujic A, Ristic M (2020) Multilevel split of high-dimensional water quality data using artificial neural networks for the prediction of dissolved oxygen in the Danube River. Neural Comput Appl 32:3957–3966
Li S, Xie G, Ren J, Yang Y, Xu X (2020) Urban PM2.5 concentration prediction via attention-based CNN-LSTM. Appl Sci 10(6):1953
Zhao Y (2020) Spatial-temporal correlation-based ISTM algorithm and its application in PM2.5 prediction. Revue Intel Artif 34(1):29–38
Aquino G, Zacarias A, Rubio JDJ, Pacheco J, Gutierrez GJ, Ochoa G, Balcazar R, Cruz DR, Garcia E, Novoa JF (2020) Novel nonlinear hypothesis for the delta parallel robot modeling. IEEE Access 8(1):46324–46334
Rubio JDJ (2009) SOFMLS: online self-organizing fuzzy modified least-squares network. IEEE Trans Fuz Syst 17(6):1296–1309
Chiang H-S, Chen M-Y, Huang Y-J (2019) Wavelet-based EEG processing for epilepsy detection using fuzzy entropy and associative Petri net. IEEE Access 7:103255–103262
Rubio JDJ (2020) Stability analysis of the modified Levenberg–Marquardt algorithm for the artificial neural network training. IEEE Trans Neural Netw Learn Syst. https://doi.org/10.1109/TNNLS.2020.3015200
Meda-Campana JA (2018) On the estimation and control of nonlinear systems with parametric uncertainties and noisy outputs. IEEE Access 6:31968–31973
Hernandez G, Zamora E, Sossa H, Tellez G, Furlan F (2020) Hybrid neural networks for big data classification. Neurocomputing 390:327–340
Raissi M (2018) Deep hidden physics models: deep learning of nonlinear partial differential equations. J Mach Learn Res 19:1–24
Lipton ZC (2018) The mythos of model interpretability. ACM Queue 16(3):1–27
Itani S, Lecron F, Fortemps P (2019) Specifics of medical data mining for diagnosis aid: a survey. Exp Sys Appl 118:300–314
Colace F, Loia V, Tomasiello S (2019) Revising recurrent neural networks from a granular perspective. Appl Soft Comput 82:105535
Loia V, Parente D, Pedrycz W, Tomasiello S (2018) A granular functional network with delay: some dynamical properties and application to the sign prediction in social networks. Neurocomputing 321:61–1
Ying H (1998) General Takagi–Sugeno fuzzy systems with simplified linear rule consequent are universal controllers, models and filters. Inf Sci 108:91–107
Bertone AM, Motta JR, Carvalho de Barros L, Gomide F (2017) Granular approximation of solutions of partial differential equations with fuzzy parameter. Gran Comput in press
Silveira G, Barros L (2015) Analysis of the dengue risk by means of a Takagi–Sugeno-style model. Fuz Sets Syst 277(15):122–137
Chen B-S, Chang Y-T (2009) Fuzzy state-space modeling and robust observer-based control design for nonlinear partial differential systems. IEEE Trans Fuz Syst 17(5):1025–1043
Matlab documentation www.mathworks.com/help/deeplearning/examples/ sequence-to-sequence-regression-using-deep-learning.html? searchHighlight=constant&s\_tid=doc\_srchtitle (accessed June 2020)
Kiselak J, Lu Y, Svihra J, Szepe P, Stehlık M (2020) “SPOCU”: scaled polynomial constant unit activation function. Neural Comput Appl (in press)
Zhu M, Min W, Wang Q, Zou S, Chen X PFLU and FPFLU: two novel non-monotonic activation functions in convolutional neural networks. Neurocomputing, 429, 110-117
Regulation (EU) 2016/679 on the protection of natural persons with regard to the processing of personal data and on the free movement of such data, and repealing. Directive 95/46/EC (General Data Protection Regulation) [2016] OJ L119/1
Bede B, Rudas IJ (2011) Approximation properties of fuzzy transforms. Fuz Sets Syst 180:20–40
Lyche T, Schumaker LL (1975) Local spline approximation methods. J Approx Theory 15(4):294–325
Pedrycz W, Vukovich W (2001) Granular neural networks. Neurocomputing 36:205–224. https://doi.org/10.1007/s40815-020-00903-z
Bellman R, Casti J (1971) Differential differential quadrature and long-term integration. J Math Anal Appl 34:235–238
Jaeger H (2001) The echo state approach to analysing and training recurrent neural networks. German National Research Center Information Technology,St. Augustin, Germany, Technical Report 148
Ryan TP (1997) Modern regression methods. Wiley, New York
Rousseeuw PJ, Leroy A (1987) Robust regression and outlier detection. Wiley, New York
Suykens JAK, De Brabanter J, Lukas L, Vandewalle J (2002) Neurocomputing 48:85–105
Ding S, Ma G, Shi Z (2014) A rough RBF neural network based on weighted regularized extreme learning machine. Neural Process Lett 40:245–260
Merikoski JK, Urpala U, Virtanen A (1997) A best upper bound for the 2-norm condition number of a matrix. Lin Alg Appl 254:355–365
Jin L, Nikifork N, Gupta MM (1994) Absolute stability conditions for discrete-time neural networks. IEEE Trans Neural Netw 5:954–964
Devaney RL (1989) An introduction to chaotic dynamical systems. Addison-Wesley, Reading
Horn RA, Johnson CA (1985) Matrix analysis. Cambridge University Press, Cambridge
Kiselak J, Lu Y, Svihra J, Szepe P, Stehlık M (2020) Correction to: “SPOCU”: scaled polynomial constant unit activation function. Neural Comput Appl
Buyukasık SA, Pashaev OK (2013) Exact solutions of forced Burgers equations with time variable coefficients. Commun Nonlin Sci Num Sim 18:1635–1651
https://www.bco-dmo.org/dataset-deployment/455062 (accessed June 2020)
Zhang S, Guo B, Dong A, He J, Xu Z, Chen SX (2017) Cautionary tales on air-quality improvement in Beijing. Proceed R Soc A 473(2205):20170457
https://archive.ics.uci.edu/ml/datasets/Beijing+Multi-Site+Air-Quality+Data Accessed June 2020
http://climate.weatheroffice.gc.ca/prods_servs/normals_documentation_e.html Accessed June 2020
Acknowledgements
Dr. Stefania Tomasiello acknowledges support from the IT Academy programme.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
The authors declare that they have no conflict of interest to this work.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
About this article
Cite this article
Tomasiello, S., Loia, V. & Khaliq, A. A granular recurrent neural network for multiple time series prediction. Neural Comput & Applic 33, 10293–10310 (2021). https://doi.org/10.1007/s00521-021-05791-4
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00521-021-05791-4