Seminario - Spatial functional data analysis: irregular spacing and Bernstein polynomials

 Abstract: Spatial Functional Data (SFD) analysis combines Functional Data Analysis (FDA) with spatial dependency modeling, treating data as continuous functions rather than scalar values or vectors. This approach offers a deeper understanding of data behavior and variability, making it ideal for analyzing continuous data collected over time, space, or other domains. Recent research introduces functional Gaussian models with Bernstein polynomial (BP) basis functions and a Bayesian framework for parameter estimation. The study assesses the BP model's ability to capture complex patterns while maintaining numerical stability and presents applications in various fields, including a practical illustration using temperature data from Mexico City.