Abstract:
|
Thousands of users rely on Stan for statistical modeling, data analysis, and prediction in the social, biological, and physical sciences, engineering, and business. Stan is a probabilistic programming language for expressing essentially any empirical model that is a differentiable function of the unknown parameters, which can then be estimated using one of the algorithms in the Stan Library. In particular, the Stan Library includes the most advanced implementation of Hamiltonian Monte Carlo, which allows researchers to efficiently draw from the posterior distribution of the unknown parameters given the known data. Alternatively, the mode of the posterior distribution can be found using conventional optimization algorithms. This talk will focus on the R interface to Stan and demonstrate how several popular regression models can be estimated using pre-written Stan code and will briefly outline how more complicated econometric models could be written using the Stan language.
|