M. Oviedo de la Fuente, V. Teodoro, C. J. Escudero, M. Febrero-Bande
Indoor air quality, of which Carbon dioxide ($CO_2$) concentration is a good indicator, plays an essential role in modern society, as we tend to spend most of our time in this kind of space. Although sometimes overlooked, it is a non-negligible factor in classrooms as it is known that poor indoor air quality is linked to several cognitive function issues, may cause loss of attention focus, and is also a virus transmission path. In this work, we will address the forecast of future values of $CO_2$ concentration levels in a classroom employing several models of the Functional Data Analysis framework such as additive, concurrent, and Hilbertian autoregressive model. These techniques will allow us to forecast the complete $CO_2$ concentration function instead of mere point forecasts. For this purpose, we will employ two datasets consisting of measurements collected in several classrooms of the University of A Coruña during the EBAU and the academic year. We will use the fda.usc package.
Keywords: Indoor air quality, $CO_2$ concentration forecast, Functional regression
Scheduled
GT20 Software and Computing for Statistics and Operations Research (SOCEIO)
June 9, 2022 12:00 PM
Grade Hall
