New Details

Speaker: Ana Martins joined IEETA in 2018 with a research grant to develop her skills in statistical modelling and data analysis. Since 2019, she is pursuing the PhD degree in Applied Mathematics (MAP-PDMA) with an individual FCT funding (Ref. SFRH/BD/143973/2019, Mathematics). Ana Martins holds a bachelor’s degree in Biochemistry (2014, University of Porto), a master’s degree in Public Health (2016, University of Porto) and a master’s degree in Mathematics and Applications (2019, University of Aveiro). Her current research interests are on time series models and spatio-temporal models, focusing the characterization of integer-valued processes.

Abstract: This work introduces the novel Space-Time Integer AutoRegressive and Moving Average (STINARMA) class of models. This class specifically accounts for the temporal and spatial dependency of integer-valued processes and is inspired by key ideas of the continuous Space-Time ARMA (STARMA) and the Integer-valued ARMA (INARMA) models. To ensure the discrete nature of the process, the binomial thinning operator replaces the multiplication of the continuous STARMA. Furthermore, the Gaussian innovations are replaced by discrete random variables. The space-time information is introduced in the model through a neighbourhood matrix, which is embedded into the matrix binomial thinning operator. The study of the moving average (STINMA), the autoregressive (STINAR) and the full STINARMA class is conducted in detail. To this purpose, the theoretical moments up to second-order, including the space-time autocovariance and autocorrelation functions, are derived. Moreover, parameter estimation is also covered in this work, namely through the Method of Moments, Conditional Least Squares (CLS) and Conditional Maximum Likelihood (CML) strategies.

Location and date: IEETA auditorium, 4th January 2024, 14h30