Non parametric functional predictive model in functional time series. Application in meteorological variables


  • Antonio Meneses Universidad Nacional de Chimborazo, Facultad de Ingeniería, Carrera de Ingeniería en Telecomunicaciones, Riobamba, Ecuador.
  • Lourdes Zúñiga Escuela Superior Politécnica de Chimborazo
  • José Muñoz Escuela Superior Politécnica de Chimborazo
  • Joge Lara Escuela Superior Politécnica de Chimborazo
  • Washington Acurio Autor Independiente, Ecuador.



Functional nonparametric model, functional time series, meteorological variables, wind speed


This research part of the study of the functional non-parametric model that is applied to functional time series. The objective is to stablish predictions of functional time series that are formed with the sample of the average wind speeds in each hour of the months of January to December of the year 2019. This sample was taken from the meteorological station of the Escuela Superior Politécnica of Chimborazo located in the San Juan Parish at an altitude of 4350 meters above sea level at kilometer 30 via Calpi - Guaranda in the province of Chimborazo - Ecuador. The R software was used to prepare the sample of the velocities and then use them as functional time series in the aforementioned model, then, with this model, the adjustments were obtained, the optimal window width typical of a non-parametric model, The predictions of the time series of 24 values ​​corresponding to each hour of the interval from 0:00 to 23:00 hours were made, which was very close to the series of wind speeds for the month of December taken as the witness month. This proximity is calculated with the mean square error MSE, less than 3%, and with wild bootstrap confidence intervals at 95%, which give the guideline to testify that the fitted model in this research is significantly reliable and it opens the way to perform future applications in other meteorological variables.


Download data is not yet available.


onparametric approach. Journal of Forecasting. 2011; 30:377-392.

Aneiros-Pérez G, Cao R, Vilar-Fernández JM. Functional Prediction for the Residual Demand in Electricity Spot Markets . IEEE Transactions on Power Systems. 2013; volumen 28.

Febrero-Bande M, Oviedo de la Fuente M. Statistical Computing in Functional Data Analysis: The R Package fda.usc. J Stat Softw. 2012.

Febrero-Bande M, Galeano P, Gozalez-Manteiga W. Functional principal component regression and functional partial least-squares regression: An overview and a comparative study. International Statistical Review. 2017; 85:61-83.

Ferraty F, Mas A, Vieu P. Nonparametric regression on functional data: inference and practical aspects. Aust. N.Z.J. Stat. 2016; 49:267-286.

Ferraty F, Keilengom I, Vieu P. On the Validity of the Bootstrap in Non Parametric Function. (2010). Springer.

Ferraty F, Vieu P. Nonparametric Functional Data Analysis. (2006). Springer.

Kokoszka P, Horv´ath L. Inference for Functional Data with Applications. Springer. 2012.

Peña D. Análisis de series de temporales. Alianza Editorial. 2010.

Portela J. Functional time series forecasting in electricity markets: a novel parametric approach. Tesis, Universidad Pontificia Comillas. 2017.

Ramsay JO, Silverman BW. Functional Data Analysis. Springer. 2005.



How to Cite

Meneses, A., Zúñiga, L., Muñoz, J., Lara, J., & Acurio, W. (2022). Non parametric functional predictive model in functional time series. Application in meteorological variables. Perfiles, 1(28).