Yusniar, Alvinda Nisma Yusniar (2020) Kajian Estimasi Parameter untuk Regresi Zero-Inlated Poisson (ZIP). Other thesis, Institut Teknologi Sepuluh Nopember.
Preview |
Text
06111540000003_Undergraduate_Thesis.pdf Download (1MB) | Preview |
Abstract
Regresi Poisson merupakan salah satu analisis regresi yang sering digunakan untuk menganalisis hubungan antara beberapa variabel respon dan variabel prediktor yang berbentuk diskrit. Akan tetapi dalam beberapa kasus terdapat fenomena dimana variabel respon yang berditribusi Poisson mengandung banyak nilai nol dengan presentase di atas 50%, fenomena tersebut dapat mengakibatkan kesalahan analisis. Oleh sebab itu diperlukan metode untuk mengatasi permasalahan tersebut.
Regresi Zero-Inflated Poisson adalah salah satu metode alternative untuk mengatasi permasalahan dimana variabel respon mengandung banyak nilai nol. Setelah dilakukan pengkajian estimasi untuk Regresi Poisson dan Regresi Zero-Infalted Poisson dengan metode Maximum Likelihood Estimation dan Moment Estimation diperoleh perbedaan langkah untuk mendapatkan nilai estimasi dan jumlah parameter. Pada regresi Poisson terdapat satu parameter sedangkan pada Regresi Poisson dan Regresi Zero-Inflated Poisson terdapat dua parameter. Langkah dalam mendapatkan nilai estimasi dengan metode Maximum Likelihood Estimation pada model Regresi Zero-Inflated Poisson membutuhkan Algoritma-EM karena hasil estimasi berbentuk implisit.
==================================================================================================================================
Poisson Regression is a regression analysis that is often used to analyze the relationship between several response variables and predictor variables that are discrete. However, in some cases there are phenomena where the response variable with Poisson distribution contains a lot of zero values with a percentage above 50%, this phenomenon can lead to analysis errors. Therefore we need a method to overcome these problems. Zero-Inflated Poisson Regression is an alternative method to solve problems where the response variable contains many zero values. After estimating the Poisson Regression and Zero-Inflated Poisson Regression with the Maximum Likelihood Estimation and Moment Estimation methods, different steps are obtained to obtain the estimated value and the number of parameters. In Poisson regression there are one parameter while in Poisson Regression and Zero-Inflated Poisson Regression there are two parameters. The step in getting the estimated value using the Maximum Likelihood Estimation method in the Zero-Inflated Poisson Regression model requires an EM-algorithm because the estimation results are in the form of implicit.
| Item Type: | Thesis (Other) |
|---|---|
| Additional Information: | RSMa 519.536 Yus k-1 2020 |
| Uncontrolled Keywords: | Regresi Poisson, Regresi Zero-Inflated Poisson, Maximum Likelihood Estimation, Moment Estimation |
| Subjects: | Q Science > QA Mathematics > QA278.2 Regression Analysis. Logistic regression |
| Divisions: | Faculty of Science and Data Analytics (SCIENTICS) > Mathematics > 44201-(S1) Undergraduate Thesis |
| Depositing User: | Alvinda Nisma Yusniar |
| Date Deposited: | 12 Mar 2025 06:09 |
| Last Modified: | 12 Mar 2025 06:09 |
| URI: | http://repository.its.ac.id/id/eprint/74596 |
Actions (login required)
 |
View Item |