Poster abstract details

The large incoming flux of asteroseismic data (but also data from spectroscopy, interferometry,...) shall lead to a better understanding of stellar interiors, their strucure and their evolution. It provides us with strong constraints on the already existing stellar models. I focus in this presentation on the interface between data analysis and stellar modelling, and address the problem of parameter estimation. I describe briefly existing methods in Bayesian analysis, and more specifically the Markov Chain Monte Carlo numerical method(s). In a second part I present results for star with available seismic data (obtained from space or ground). An emphasis is put on the necessity to adapt the methodology (tempered MCMC, model selection method, inclusion of interpolation schemes) to the specific problem of stellar physics that is addressed.