PERBANDINGAN METODE HOLT WINTER’S EXPONENTIAL SMOOTHING DAN EXTREME LEARNING MACHINE UNTUK PERAMALAN JUMLAH BARANG YANG DIMUAT PADA PENERBANGAN DOMESTIK DI BANDARA UTAMA SOEKARNO HATTA

Abstract

The loading of goods carried out at the airport is an essential part of the transporting goods system. In this regard, it is necessary to have a prediction to make the right policy or to solve the problems that occur. Holt Winter's Exponential Smoothing, which one of the classic methods of analyzing time series data, and Extreme Learning Machine which is part of the artificial neural network method, are methods that can be used as a tool for forecasting problems. Holt Winter's Exponential Smoothing uses three times of smoothing on related data, which are level smoothing, trend smoothing, and season smoothing, while Extreme Learning Machine goes through three stages, which are normalization, training, and denormalization. In measuring the error rate in related forecasting, the symmetric Mean Absolute Percentage Error (sMAPE) value is used. The Holt Winter's Exponential Smoothing method Additive model produces a sMAPE value of 26.14%; while the Multiplicative model with the same method resulted in the sMAPE value of 25.69%. For the Extreme Learning Machine method, the sMAPE value is 49.85%. Based on the accuracy test using the sMAPE value, Holt Winter's Exponential Smoothing method Multiplicative model is the better method than Extreme Learning Machine