The Partial Autocorrelation Function Analysis in Predicting Speed Wind Maximum Georgina Maria Tinungki
Department of Mathematics
Faculty of Mathematics and Natural Science
Hasanuddin University
Makassar 90245, Indonesia
Abstract
The stationary and non-stationary examination of the data set can be done by plot analysis of the Partial Autocorrelation Function of the data, by looking at the maximum number of Partial Autocorrection Function Fixed value estimates. The Autocorrelation Function (ACF) is a function that shows the magnitude of the correlation between the observation of the t-time and the observation at the previous time. The autocorrelation function shows the autocorrelation coefficient which is the measurement of correlation between the observations at different times. Data taken from Central Bureau of Statistics of Makassar City, Indonesia is data about the maximum wind speed by month at Paotere station in Makassar City, Indonesia in years 2009 - 2016. From this data will predict the maximum wind speed during the next 24 months, ie from January 2017 to December 2018. The results obtained, forecasting is done with 12 leads period ahead with 95% confidence interval.
Keywords: Partial Autocorrelation Function, stationary and no stationer data, Autocorrelation Function
