Seminar: Statistical prediction in variables related to energy production will take place on June, 29.
On June 29, 2017 at 11:00 hours the Seminar: Statistical prediction in variables related to energy production will take place at the Salón de Grados of the Faculty of Mathematics of the University of Santiago de Compostela. (USC).
The seminar is structured in 2 talks. The first one, given by Manuel Oviedo de la Fuente, ITMATI Contracted Researcher and titled: "Selection of variables applied to the prediction of demand and the price of energy." The second of them, will be taught by Jairo Cugliari, University of Lyon 2, entitled "Non parametric forecasting and functional clustering using wavelets. Application to electricity demand ".
PROGRAM:
- 12:00 – 13:00 Manuel Oviedo de la Fuente (University of Santiago de Compostela e ITMATI) “Selection of variables applied to the prediction of demand and the price of energy."
Summary:
This work considers the problem of variable selection when some of the variables have a functional nature (such as the daily energy price curve) that are related to other variables (scalar, multivariate, directional, etc.). Our proposal starts from a null regression model and sequentially selects a new variable that will be incorporated into the model until all relevant and non-redundant information is included. For this, a comprehensive use of the distance correlation, R, proposed by Szekely et al. (2007) that allows to quantify the dependence between two variables of arbitrarily finite dimensions. The procedure has shown promising results when applied to actual data sets. As an illustration, the proposal has been applied to forecast the demand and price of energy in the Iberian market using information from different sources such as the type of energy generated (nuclear, coal, combined cycle, renewable, etc.), meteorological data (Temperature, solar radiation, wind speed, etc.) and seasonal variables (year, month, day of the week, holidays, etc.).
This work considers the problem of variable selection when some of the variables have a functional nature (such as the daily energy price curve) that are related to other variables (scalar, multivariate, directional, etc.). Our proposal starts from a null regression model and sequentially selects a new variable that will be incorporated into the model until all relevant and non-redundant information is included. For this, a comprehensive use of the distance correlation, R, proposed by Szekely et al. (2007) that allows to quantify the dependence between two variables of arbitrarily finite dimensions. The procedure has shown promising results when applied to actual data sets. As an illustration, the proposal has been applied to forecast the demand and price of energy in the Iberian market using information from different sources such as the type of energy generated (nuclear, coal, combined cycle, renewable, etc.), meteorological data (Temperature, solar radiation, wind speed, etc.) and seasonal variables (year, month, day of the week, holidays, etc.).
- 13:00 – 14:00 Jairo Cugliari (University of Lyon 2) “Non parametric forecasting and functional clustering using wavelets. Application to electricity demand”
Summary:
This talk has an industrial motivation that is the nonparametric forecast of electricity demand for the French producer EDF. We then present two methods for detecting patterns and clusters in high dimensional time-dependent functional data. Our methods are based on wavelet-based similarity measures, since wavelets are well suited for identifying highly discriminant time-scale features. The multiresolution aspect of the wavelet transform provides a time-scale decomposition of the signals allowing to visualize and to cluster the functional data into homogeneous groups. For each input function, through its empirical orthogonal wavelet transform the first method uses the distribution of energy across scales to generate a representation that can be sufficient to make the signals well distinguishable. Our new similarity measure combined with a feature selection technique is then used within classical clustering algorithms to effectively differentiate among high dimensional populations. The second method uses similarity measures between the whole time-scale representations that are based on wavelet-coherence tools. The clustering is then performed using a k-centroid algorithm starting from these similarities. Finally the practical performance of these methods is illustrated through the daily profiles of the French electricity power demand involved in nonparametric forecasting as well as individual consumers clustering involved in the forecasting by disaggregation of the electricity consumption. The talk is related to joint works with Anestis Antoniadis, Xavier Brossat, Yannig Goude and Jean-Michel Poggi.
This talk has an industrial motivation that is the nonparametric forecast of electricity demand for the French producer EDF. We then present two methods for detecting patterns and clusters in high dimensional time-dependent functional data. Our methods are based on wavelet-based similarity measures, since wavelets are well suited for identifying highly discriminant time-scale features. The multiresolution aspect of the wavelet transform provides a time-scale decomposition of the signals allowing to visualize and to cluster the functional data into homogeneous groups. For each input function, through its empirical orthogonal wavelet transform the first method uses the distribution of energy across scales to generate a representation that can be sufficient to make the signals well distinguishable. Our new similarity measure combined with a feature selection technique is then used within classical clustering algorithms to effectively differentiate among high dimensional populations. The second method uses similarity measures between the whole time-scale representations that are based on wavelet-coherence tools. The clustering is then performed using a k-centroid algorithm starting from these similarities. Finally the practical performance of these methods is illustrated through the daily profiles of the French electricity power demand involved in nonparametric forecasting as well as individual consumers clustering involved in the forecasting by disaggregation of the electricity consumption. The talk is related to joint works with Anestis Antoniadis, Xavier Brossat, Yannig Goude and Jean-Michel Poggi.
REGISTRATION: Registration is free: Registration form (click here)
A certificate of attendance will be issued for those who request it.
Consellería de Cultura, Educación e Ordenación Universitaria of the Xunta de Galicia collaborates with this seminar through the Technological Network of Industrial Mathematics (TMATI Network) and the agreement that ITMATI has with this Consellería.
Fecha:
Thu, 2017-06-15 
