Dynamic Quantile Linear Models: A Bayesian Approach
Year: 2019
Type: article
Abstract: The paper introduces a new class of models, named dynamic quantile linear models, which combines dynamic linear models with distribution-free quantile regression producing a robust statistical method. Bayesian estimation for the dynamic quantile linear model is performed using an efficient Markov chain Monte Carlo algorithm. The paper also proposes a fast sequential procedure suited for high-dimensional predictive modeling with massive data, where the generating process is changing over time. The proposed model is evaluated using synthetic and well-known time series data. The model is also applied to predict annual incidence of tuberculosis in the state of Rio de Janeiro and compared with global targets set by the World Health Organization.
Source: Bayesian Analysis
DOI: 10.1214/19-ba1156
Institutions: Universidade Federal do Rio de Janeiro, Fundação Oswaldo Cruz
Language: English
Subfield: Statistics and Probability
Field: Mathematics
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W2964247004
