This Author published in this journals
All Journal Media Statistika
Winita Sulandari
Unknown Affiliation

Published : 3 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 3 Documents
Search

ESTIMASI PARAMETER MODEL MIXTURE AUTOREGRESSIVE (MAR) MENGGUNAKAN ALGORITMA EKSPEKTASI MAKSIMISASI (EM) Asrini, Mika; Sulandari, Winita; Wiyono, Santoso Budi
MEDIA STATISTIKA Vol 6, No 1 (2013): Media Statistika
Publisher : Department of Statistics, Faculty of Science and Mathematics, Universitas Diponegoro

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (268.444 KB) | DOI: 10.14710/medstat.6.1.21-26

Abstract

Mixture autoregressive (MAR) Model is a mixture of Gaussian autoregressive (AR) components. The mixture model is capable for modelling of nonlinear time series with multimodal conditional distributions. This paper discusses about the parameters estimation using EM algorithm. All possible models are then applied to national maize production data. In this case, the BIC is used for the MAR model selection. Keywords : Mixture Autoregressive, EM Algorithm, BIC, Maize Production
PERAMALAN PENGGUNAAN BEBAN LISTRIK JANGKA PENDEK GARDU INDUK BAWEN DENGAN DSARIMA Saptyani, Marita; Sulandari, Winita; Pangadi, Pangadi
MEDIA STATISTIKA Vol 8, No 1 (2015): Media Statistika
Publisher : Department of Statistics, Faculty of Science and Mathematics, Universitas Diponegoro

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (541.279 KB) | DOI: 10.14710/medstat.8.1.41-48

Abstract

Bawen substation is a part of electrical distribution system. Forecasting load demand is required for power planning. Data used in this research are an hourly load demand of Bawen, Salatiga for 3 months, from February 2, 2013 to April 29, 2013, measured in Megawatt (MW).A half hourly load demand forecasting is needed for real time controlling and short-term maintenance schedulling. Since the data have two seasonal periods, i.e. daily and weekly seasonality with length 48 and 336 respectively, the model of double seasonal ARIMA (DSARIMA) is proposed as the most appropriate model for the case. Initial model is determined by the pattern of the data, based on the autocorrelation function plot. Some experiments was done by choosing several periods data. The most suitable model is chosen based on the outsample mean absolute percentage error (MAPE). The current study shows that the DSARIMA (0, 1, [1, 20, 47])(0, 1, 1)48(0, 1, 0)336 is the best model to forecast  336 next period. Keywords: DSARIMA, MAPE, Electricity, Bawen
PENERAPAN MODEL HYBRID ARIMA BACKPROPAGATION UNTUK PERAMALAN HARGA GABAH INDONESIA Janah, Sufia Nur; Sulandari, Winita; Wiyono, Santoso Budi
MEDIA STATISTIKA Vol 7, No 2 (2014): Media Statistika
Publisher : Department of Statistics, Faculty of Science and Mathematics, Universitas Diponegoro

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (591.422 KB) | DOI: 10.14710/medstat.7.2.63-69

Abstract

Hybrid model discussed in this paper combining ARIMA and backpropagation is applied to grain price forecasting in Indonesia for period January 2008 until April 2013. The grain price time series consists of linear and nonlinear patterns. Backpropagations can recognize non linear patterns that can not be done by ARIMA. In order to find the best model, some combinations of prepocessing transformations, the number of input and hidden units, and the activation function were applied in the contruction of the network structure. Based on the experiments, it can be showed that ARIMA backpropagation hybrid model provides more accurate results than ARIMA model.  The hybrid model would rather be used in the short-term forecasting, no more than three periods. Keywords: ARIMA, Backpropagation, Hybrid, Grain Price