Bayesian predictive synthesis (BPS) is a foundational framework for evaluation, calibration, comparison, and context-and data-informed combination of multiple forecast distributions arising from multiple models or sources. BPS subsumes and explains existing density forecasts combination methods, helps to identify strengths and weaknesses of specific approaches, and defines forecast synthesis when predictions come from sets of models, individual forecasters, agencies, or other sources. My talk highlights foundational aspects and then applications of BPS to multi-step ahead macro-economic forecasting. The latter involves BPS based on multivariate, dynamic latent factor models in which latent factor processes represent individual models or forecasters. The framework allows modelling and estimation-- sequentially and adaptively over time-- of varying forecast biases and facets of miscalibration of individual forecast densities, and of time-varying inter-dependencies among models or forecasters over multiple series.
Texto informado pelo autor.
Registre-se com antecedência para este seminário:
Após o registro, você receberá um e-mail de confirmação contendo informações sobre conexão no seminário.
Mike West is the Arts & Sciences Professor of Statistics & Decision Sciences the Department of Statistical Science at Duke University. Mike's research and teaching activities are in Bayesian analysis in ranges of interlinked areas: theory and methods of dynamic models in time series analysis, multivariate analysis, latent structure, high-dimensional inference and computation, quantitative and computational decision analysis, stochastic computational methods, and statistical computing, among other topics.