Search

This paper describes the R package VSURF. Based on random forests, and for both regression and classification problems, it returns two subsets of variables. The first is a subset of important variables including some redundancy which can be relevant for interpretation, and the second one is a smaller subset corresponding to a model trying to avoid redundancy focusing more closely on the prediction objective. The two-stage strategy is based on a preliminary ranking of the explanatory variables using the random forests permutation-based score of importance and proceeds using a stepwise forward strategy for variable introduction. The two proposals can be obtained automatically using data-driven default values, good enough to provide interesting results, but strategy can also be tuned by the user. The algorithm is illustrated on a simulated example and its applications to real datasets are presented. [ABSTRACT FROM AUTHOR]

The authors deal with forecasting nonstationary time series using wavelets and kernel smoothing. Starting from a basic forecasting procedure based on the regression of the process on the nondecimated Haar wavelet coefficients of the past, the procedure was extended in various directions, including the use of an arbitrary wavelet or polynomial fitting for extrapolating low-frequency components. The authors study a further generalization of the prediction procedure dealing with multistep forecasting and combining kernel smoothing and wavelets. They finally illustrate the proposed procedure on nonstationary simulated and real data and then compare it to well-known competitors. [ABSTRACT FROM AUTHOR]

Abstract: In this paper, we are interested in the statistical forecasting of the daily mean PM10 concentration. Hourly concentrations of PM10 have been measured in the city of Rouen, in Haute-Normandie, France. Located at northwest of Paris, near the south side of Manche sea and heavily industrialised. We consider three monitoring stations reflecting the diversity of situations: an urban background station, a traffic station and an industrial station near the cereal harbour of Rouen. We have focused our attention on data for the months that register higher values, from December to March, on years 2004–2009. The models are obtained from the winter days of the four seasons 2004/2005 to 2007/2008 (training data) and then the forecasting performance is evaluated on the winter days of the season 2008/2009 (test data). We show that it is possible to accurately forecast the daily mean concentration by fitting a function of meteorological predictors and the average concentration measured on the previous day. The values of observed meteorological variables are used for fitting the models and are also considered for the test data. We have compared the forecasts produced by three different methods: persistence, generalized additive nonlinear models and clusterwise linear regression models. This last method gives very impressive results and the end of the paper tries to analyze the reasons of such a good behavior. [Copyright &y& Elsevier]

Abstract: This paper proposes, focusing on random forests, the increasingly used statistical method for classification and regression problems introduced by Leo Breiman in 2001, to investigate two classical issues of variable selection. The first one is to find important variables for interpretation and the second one is more restrictive and try to design a good parsimonious prediction model. The main contribution is twofold: to provide some experimental insights about the behavior of the variable importance index based on random forests and to propose a strategy involving a ranking of explanatory variables using the random forests score of importance and a stepwise ascending variable introduction strategy. [Copyright &y& Elsevier]

Forecasting PM10 concentrations accurately will all for improved early warning procedures, useful for safety reasons and opens for example the possibility to restrict circulation or to decide free public transportation. So the need of a statistical pollution forecasting tool from particulate matter is an important issue for the public authorities. Hourly concentrations of PM10 have been measured in three cities of Haute-Normandie (France): Rouen, Le Havre and Dieppe. The Haute-Normandie region is located at northwest of Paris, near the south side of Manche sea and is heavily industrialized. We consider six monitoring stations reflecting the diversity of situations. We have focused our attention on recent data from 2007 to 2011. We forecast the daily mean PM10 concentration by modeling it as a mixture of linear regression models involving meteorological predictors and the average concentration measured on the previous day. The values of observed meteorological variables are used for fitting the models while the corresponding predictions are considered for the test data, leading to realistic evaluations of forecasting performances, which are calculated through a leave-one-out scheme on the four years. We discuss in this paper several methodological issues including estimation schemes, introduction of the deterministic predictions of meteorological models and how to handle the forecasting at various horizons from some hours to one day ahead. [ABSTRACT FROM AUTHOR]