Bayesian concurrent functional regression for sparse data.

• Beatrice Charamba ^[1] ; Andrew J Simpkin ^[1]
  1. [1] National University of Ireland

     National University of Ireland

     Irlanda

     
• Localización: Proceedings of the 35th International Workshop on Statistical Modelling : July 20-24, 2020 Bilbao, Basque Country, Spain / Itziar Irigoyen Garbizu (ed. lit.), Dae-Jin Lee (ed. lit.), Joaquín Martínez Minaya (ed. lit.), María Xosé Rodríguez Álvarez (ed. lit.), 2020, ISBN 978-84-1319-267-3, págs. 45-50
• Idioma: inglés
• Enlaces
  • Texto Completo Libro
• Resumen

  • The recent abundance of wearable technology has led to a sharp rise in the availability of multivariate data streams. However, many functional data analysis (FDA) methods require such data to be measured regularly without missingness, with data being collected at the same fixed times for all individuals.

    In order to deal with irregular, concurrent, functional data including missing values, we developed the Bayesian model for function-on-function regression. This method is tested in a simulation study and applied to concurrently measured glucose (every 5 minutes for 1 week) and electrocardiogram (ECG) data (every 10 minutes for 1 week) in a cohort of n = 17 type 1 diabetics. The Bayesian model outperformed other models when the underlying relationship is complex and non-linear