All Times EDT
Keywords: Time Series Analysis, Functional Data Analysis, ARMA, PM2,5 stationarity, Hilbert Spaces
Time series models are one of the most popular methodologies for explaining the behavior of data that is temporarily indexed. Recently, the functional data time series is a growing field with continuous methodological developments. In this work, we propose an adaptation of autoregressive-moving average (ARMA) functional model (FARMA) for PM2.5 data recorded in Cali, Colombia during 2018. The main advantage of FARMA compared to the conventional ARMA model was that the adjusting of this model did not require to manipulate the outliers points and the dimensional of the data was not reduced.