Darussalam, Indra Alim (2019) Analisis Faktor yang Mempengaruhi Jumlah Kejadian Puting Beliung Indonesia Menggunakan Regresi Poisson. Other thesis, Institut Teknologi Sepuluh Nopember.
Preview |
Text
06111440000096-undergraduate_thesis.pdf Download (1MB) | Preview |
Abstract
Model regresi Poisson merupakan salah satu model regresi yang dapat menunjukkan hubungan antara variabel respon berdistribusi Poisson dengan variabel-variabel prediktor melalui parameter regresi. Regresi Poisson umumnya digunakan untuk menganalisis data dengan asumsi nilai rata-rata dan varian sama (equidispersi). Akan tetapi, secara umum data sering kali terjadi nilai varian melebihi rata-rata (overdispersi). Generalized Poisson Regression (GPR) digunakan untuk mengatasi masalah tersebut. Pada Tugas Akhir ini membahas model regresi Poisson dan GPR kejadian puting beliung di Indonesia tahun 2015, 2016 dan 2017. Penaksiran Parameter model regresi Poisson dan GPR menggunakan metode Maximum Likelihood Estimation (MLE) dan iterasi Newton-Raphson. Berdasarkan pengolahan data diketahui model regresi Poisson terjadi overdispersi dan dilakukan Generalized Poisson Regression untuk mengatasi masalah tersebut. Model GPR puting beliung 2015 adalah \lambda_i=exp(1,2319+1,3903x_{2i}) yang menunjukan rata-rata jumlah kejadian puting beliung pada pengamatan provinsi ke-i adalah exp(1,2319) dengan bertambahnya exp(1,3903) kecepatan angin. Model GPR puting beliung 2016 adalah adalah \lambda_i=exp(19,5612-4,2141x_{2i}+0,268x_{3i}-0,05745x_{6i}) yang menunjukan rata-rata jumlah kejadian puting beliung pada pengamatan provinsi ke-i adalah exp(19,5612) dengan bertambahnya exp(4,2141) kecepatan angin, bertambahnya exp(0,268) kelembaban udara, berkurangnya exp(0,05745) penyinaran matahari. Model GPR puting beliung 2017 adalah adalah \lambda_i=exp(27,1462-0,2946x_{3i}) yang menunjukan rata-rata jumlah kejadian puting beliung pada pengamatan provinsi ke-i adalah exp(27,1462) dengan berkurangnya exp(0,2946) kelembaban udara.
=================================================================================================================================
The Poisson regression model is one of the regression models that can show the relationship between response variables Poisson distribution and predictor variables through regression parameters. Poisson regression is generally used to analyze data with assumptions of equal and equal variance (equidispersion). However, in general, data variants often occur over average values (overdispersion). Generalized Poissson Regression (GPR) is used to overcome this problem. In this study discuss Poisson regression models and GPR tornado occurrences in Indonesia in 2015, 2016, and 2017. The parameter estimation of Poisson regression models and (GPR) using the Maximum Likelihood Estimation (MLE) method and Newton-Raphson iteration. Based on data processing, it is known that the Poisson regression model is overdispersed and Generalized Poissson Regression is performed to overcome this problem. The GPR model for tornado in 2015 was \lambda_i=exp(1,2319+1,3903x_{2i}) which showed that the average number of tornado incidents in 2015 for every i observation was exp (3.8821) with reduced exp (0,00039) rainfall. The GPR model for 2016 tornadoes is \lambda_i=exp(19,5612+4,2141x_{2i}+0,268x_{3i}-0,05745x_{6i}) which shows the average number of tornado incidents in 2016 at each i observation is exp (19,5612) with increase exp (4,2141) wind speed, increase in exp (0,268) humidity, decreased exp (0,05745) sun radiation. The GPR model for tornado in 2017 is \lambda_i=27,1462-0,2946x_{3i} which shows that the average number of tornado incidents in 2017 for each i observation is exp (27,1462) with a decrease in exp (0.2946) air humidity.
| Item Type: | Thesis (Other) |
|---|---|
| Additional Information: | RSMa 519.536 Dar a-1 2019 |
| Uncontrolled Keywords: | Puting Beliung, Maximum Likelihood Estimation, Newton-Raphson, Overdispersi, regresi Poisson, Generalized Poisson Regression |
| Subjects: | G Geography. Anthropology. Recreation > GE Environmental Sciences Q Science > QA Mathematics > QA278.2 Regression Analysis. Logistic regression |
| Divisions: | Faculty of Mathematics, Computation, and Data Science > Mathematics > 44201-(S1) Undergraduate Thesis |
| Depositing User: | Darussalam Indra Alim |
| Date Deposited: | 25 Jan 2024 04:18 |
| Last Modified: | 25 Jan 2024 04:18 |
| URI: | http://repository.its.ac.id/id/eprint/65913 |
Actions (login required)
 |
View Item |