Showing 1–3 of 3 results for author: Dacunha-Castelle, D

Aggregation of autoregressive processes and long memory

Authors: Didier Dacunha-Castelle, Lisandro J. Fermín

Abstract: We study the aggregation of AR processes and generalized Ornstein-Uhlenbeck (OU) processes. Mixture of spectral densities with random poles are the main tool. In this context, we apply our results for the aggregation of doubly stochastic interactives processes, see Dacunha-Castelle and Fermin (2006). Thus, we study the relationship between aggregation of autoregressive processes and long memory… ▽ More We study the aggregation of AR processes and generalized Ornstein-Uhlenbeck (OU) processes. Mixture of spectral densities with random poles are the main tool. In this context, we apply our results for the aggregation of doubly stochastic interactives processes, see Dacunha-Castelle and Fermin (2006). Thus, we study the relationship between aggregation of autoregressive processes and long memory considering complex interaction structures. We precise a very interesting qualitative phenomena: how the long memory creation depends on the poles concentration near to the boundary of stability (measured in the Prokhorov sense). Our results extends the results given by Oppenheim and Viano (2004), and highlight the importance of the angular dispersion measure of poles in the appearance of the long memory. △ Less

Submitted 12 November, 2008; originally announced November 2008.

Comments: 15 pages, 2 figures

MSC Class: 60G10; 60G50; 62M10; 91B84

Estimation of a diffusion model with trends taking in account the extremes. Application to temperature in France

Authors: Gregory Benmenzer, Didier Dacunha-Castelle, T. T. Huong Hoang

Abstract: We built a model of the daily temperature based on a diffusion process and address to extreme values not taken into account in the literature on this kind of models. We first study, using non parametric tools, the trends on mean and variance. In a second step we estimate a stationary model first non parametrically and then using likelihood methods. Extreme values are taken into account in the es… ▽ More We built a model of the daily temperature based on a diffusion process and address to extreme values not taken into account in the literature on this kind of models. We first study, using non parametric tools, the trends on mean and variance. In a second step we estimate a stationary model first non parametrically and then using likelihood methods. Extreme values are taken into account in the estimation of model and to obtain a definitive estimation we use in a specific framework extreme theory for diffusions. A test of suitable model by simulation is done. △ Less

Submitted 23 June, 2008; originally announced June 2008.

Disaggregation of Long Memory Processes on C^{\infty} Class

Authors: Didier Dacunha-Castelle, Lisandro Fermín

Abstract: We prove that a large set of long memory (LM) processes (including classical LM processes and all processes whose spectral densities have a countable number of singularities controlled by exponential functions) are obtained by an aggregation procedure involving short memory (SM) processes whose spectral densities are infinitely differentiable (C^{infty}). We show that the C^{infty} class of spec… ▽ More We prove that a large set of long memory (LM) processes (including classical LM processes and all processes whose spectral densities have a countable number of singularities controlled by exponential functions) are obtained by an aggregation procedure involving short memory (SM) processes whose spectral densities are infinitely differentiable (C^{infty}). We show that the C^{infty} class of spectral densities is the optimal class to get a general result for disaggregation of LM processes in SM processes, in the sense that the result given in C^{infty} class cannot be improved taking for instance analytic functions instead of indefinitely derivable functions. △ Less

Submitted 14 September, 2005; originally announced September 2005.

Comments: Submitted to Electronic Communications in Probability. 10 pages

MSC Class: 60E07; 60G10; 60G50; 62M10

Journal ref: Electronic Communications in Probability . Vol. 11(2006) . Paper 4