Fuzzy transforms and seasonal time series

Abstract

Like in our previous papers, we show the trend of seasonal time series by means of polynomial interpolation and we use the inverse fuzzy transform for prediction of the value of an assigned output. As example, we use the daily weather dataset of the city of Naples (Italy) starting from data collected from 2003 till to 2015 making predictions on the the relative humidity parameter. We compare our method with the traditional F-transform based, the average seasonal variation and the famous ARIMA methods.