Title: Consistent model selection for ergodic SDEs
Authors: Shoichi Eguchi - Osaka University (Japan) [presenting]
Yuma Uehara - The Institute of Statistical Mathematics (Japan)
Abstract: There are several studies of model selection for stochastic differential equations (SDEs), for example, the contrast-based information criterion for ergodic diffusion processes and the Schwarz type information criterion for locally asymptotically quadratic models. However, most of the existing theoretical literature has been developed for nested models. We will give the mathematical validity of Bayesian model comparison for possibly misspecified ergodic SDE models and propose the quasi-Bayesian information criterion (QBIC).