EXPONENTIAL SMOOTHING TECHNIQUES ON DAILY TEMPERATURE LEVEL DATA

The changes of temperature level occur throughout the year. This event whether hot temperature or cold temperature can affect human life and nature. Such event is also known as extreme event due to the nature of the data produced. Usually the time series of extreme dataset is rarely linear. The existence of nonlinear pattern and high fluctuation in variation greatly affect the quality of forecasting performances. Three exponential smoothing techniques have been tested to study their ability in handling of temperature level data from three cities in Texas. Single Exponential Smoothing Technique (SEST), Double Exponential Smoothing Technique (DEST) and Holt's method were explored in preparing the temperature data. From the experiments, it was found that DEST is the most suitable technique to deal with the data compared to SEST and Holt's method.