Detalles de Evento
Día y hora: Jueves 10 de Diciembre, 12:00 - 13:00
Lugar: Aula Virtual Newton
Acceso Sala Newton
- Contraseña de la reunión: 339336
Es obligatorio que todos los participantes accedan con este enlace, no use otro metodo de acceso.
Título: APPROXIMATED TRANSITION DENSITIES OF LOGISTIC-TYPE DIFFUSIONS VIA HERMITE EXPANSIONS.
Ponente: Antonio Barrera García.
Dpto. de Análisis Matemático, Estadística e Investigación Operativa y Matemática Aplicada. Universidad de Málaga
Abstract: Logistic-type dynamics are widely used in modelling different phenomena. Following classic ordinary differential equations, continuous time-inhomogeneous diffusion processes can be obtained by including a time-dependent growth rate function and noise-perturbed regulation functions. This usually leads to unknown transition densities, complicating other procedures such as likelihood estimation. Approximations of such unknown functions may help to deal with this issue, provided limit theorems guarantee the convergence to the true densities. Hermite expansion around the standard gaussian distribution have been a successful approach to the problem. Previous transformations of the original density are required in order to reduce the assumptions about normality which could not be met in some practical applications. In this work, this procedure is revised and adapted to the case of time-inhomogeneous diffusion processes with logistic mean behaviour. Some results are established in terms of the time-dependent growth rate function. Remarks about the influence of the sole growth rate function are discussed.
Lugar: Aula Virtual Newton
Acceso Sala Newton
- Contraseña de la reunión: 339336
Es obligatorio que todos los participantes accedan con este enlace, no use otro metodo de acceso.
Título: APPROXIMATED TRANSITION DENSITIES OF LOGISTIC-TYPE DIFFUSIONS VIA HERMITE EXPANSIONS.
Ponente: Antonio Barrera García.
Dpto. de Análisis Matemático, Estadística e Investigación Operativa y Matemática Aplicada. Universidad de Málaga
Abstract: Logistic-type dynamics are widely used in modelling different phenomena. Following classic ordinary differential equations, continuous time-inhomogeneous diffusion processes can be obtained by including a time-dependent growth rate function and noise-perturbed regulation functions. This usually leads to unknown transition densities, complicating other procedures such as likelihood estimation. Approximations of such unknown functions may help to deal with this issue, provided limit theorems guarantee the convergence to the true densities. Hermite expansion around the standard gaussian distribution have been a successful approach to the problem. Previous transformations of the original density are required in order to reduce the assumptions about normality which could not be met in some practical applications. In this work, this procedure is revised and adapted to the case of time-inhomogeneous diffusion processes with logistic mean behaviour. Some results are established in terms of the time-dependent growth rate function. Remarks about the influence of the sole growth rate function are discussed.
