Pramujati, Windya Harieska (2020) Aplikasi Model Binomial Negatif Generalized Seasonal Autoregressive Integrated Moving Average (GSARIMA) pada Peramalan Jumlah Penderita Penyakit Infeksi Saluran Pernapasan Akut (ISPA). Other thesis, Institut Teknologi Sepuluh Nopember.
Preview |
Text
06111640000030-Undergraduate_Theses.pdf Download (2MB) | Preview |
Abstract
Infeksi Saluran Pernapasan Akut (ISPA) merupakan penyakit yang disebabkan oleh virus atau bakteri. Jumlah penderita ISPA biasanya meningkat karena adanya pengaruh musim. Salah satu provinsi di Indonesia dengan prevalensi ISPA tertinggi adalah Jawa Timur yaitu di Kota Surabaya. Berdasarkan fakta tersebut, maka akan dilakukan peramalan jumlah penderita ISPA di Kota Surabaya. Metode peramalan yang sering digunakan pada data time series adalah ARIMA. Namun pada peramalan data count, model Gaussian tidak selalu tepat digunakan, dan pada penelitian ini ditemukan adanya overdispersion (nilai varian variabel respon lebih besar dari pada nilai mean). Sehingga diterapkan sebuah model peramalan pada data count dengan pendekatan distribusi Binomial Negatif dan melibatkan efek musiman yaitu model Binomial Negatif GSARIMA. Pada penelitian ini didapatkan model Binomial Negatif GSARIMA(2,1,1)(0,1,1)6 yang didapatkan dari identifikasi model SARIMA. Estimasi parameter model GSARIMA didapatkan dengan metode Bayesian. Terdapat dua model GSARIMA yaitu model GSARIMA transformasi ZQ1 dan transformasi ZQ2. Hasil yang didapatkan, peramalan dengan model Binomial Negatif GSARIMA transformasi ZQ1 lebih baik dibandingkan transformasi ZQ2 dengan nilai MARE sebesar 0,1311 yang diaplikasikan pada data jumlah penderita ISPA.
=========================================================
Acute Respiratory Infection (ARI) is a disease caused by a virus or bacteria. The number of ARI sufferers usually increases due to seasonal influences. One of the provinces in Indonesia with the highest ARI prevalence is East Java, that is in Surabaya city. Based on these facts, forecasting the number of ARI in Surabaya will be conducted. Forecasting method, that is often used in time series data is ARIMA. However, in forecasting data count, the Gaussian model is not always appropriate, and this research found overdispersion (the variance value of the response variable is greater than the mean value). So the forecasting model is applied to the data count with Negative Binomial distribution and involves seasonal effects, that is Negative Binomial GSARIMA model. In this research, the Negative Binomial GSARIMA model (2,1,1)(0,1,1)6 was obtained from the identification of the SARIMA model. The estimated parameters of the GSARIMA model are obtained by the Bayesian method. There are two GSARIMA models, namely the GSARIMA ZQ1 transformation model and ZQ2 transformation. The results obtained, forecasting with the Negative Binomial GSARIMA ZQ1 transformation model is better than the ZQ2 transformation with a MARE value is 0.1311 which is applied to data the number of ARI patients.
| Item Type: | Thesis (Other) |
|---|---|
| Uncontrolled Keywords: | ARI, Negative Binomial GSARIMA, SARIMA |
| Subjects: | Q Science Q Science > QA Mathematics Q Science > QA Mathematics > QA276 Mathematical statistics. Time-series analysis. Failure time data analysis. Survival analysis (Biometry) |
| Divisions: | Faculty of Science and Data Analytics (SCIENTICS) > Mathematics > 44201-(S1) Undergraduate Thesis |
| Depositing User: | Windya Harieska Pramujati |
| Date Deposited: | 21 Aug 2020 02:19 |
| Last Modified: | 06 Jul 2023 13:24 |
| URI: | http://repository.its.ac.id/id/eprint/79307 |
Actions (login required)
 |
View Item |