Ariani, Nelly (2013) Perbandinganregresi Zero Inflated Generalized Poisson (Zigp) Danregresi Zero Inflated Negative Binomial (Zinb)Pada Data Overdispersion. Sarjana thesis, Universitas Brawijaya.
Abstract
Analisisregresimerupakan metode yang digunakanuntukmengetahuihubunganketergantunganantarapeubahrespon (Y) denganpeubah prediktor (X). Regresi Poissonmerupakan analisis regresi untuk data diskrit atau count. Padaregresi Poisson terdapatasumsi nilai rata-rata yang samadengannilai ragam (equidispersion) tetapi dapat dijumpai data yang memiliki nilai ragam lebih besar dari nilai rata-ratanya(overdispersion), sehingga regresi Poisson tidak tepat lagi digunakan untuk memodelkan data. Model regresi yang lebihsesuaiuntuk data overdispersionadalah model regresi Generalized Poisson (GP)danNegative Binomial (NB). Pada peubah respon sering dijumpai adanya data yang bernilai nol dan proporsinya besar (zero inflation) yaitu lebih dari 50%. Besarnya proporsi data yang bernilai nol dapat berakibat pada ketepatan (presisi) dari inferensia. Selain itu, regresi Poissonmenjadi tidak tepat lagi memodelkan data yang sebenarnya. Alternatif model regresi yang lebihsesuaiuntuk dataoverdispersiondan zero inflationpada peubah respon adalah model regresi Zero Inflated Generalized Poisson (ZIGP)danZero Inflated Negative Binomial (ZINB). Penelitianinibertujuanuntukmembandingkan model regresi Zero InflatedGeneralized Poisson (ZIGP) danZero InflatedNegative Binomial (ZINB) berdasarkannilai AIC (Akaike Information Criterion). Data yang digunakanadalah 3 data sekunder yang overdispersiondan zero inflation pada peubah respon.Hasilanalisismenunjukkanmodel Zero Inflated Generalized Poisson (ZIGP) lebihsesuaijikadigunakanpada data overdispersiondanzero inflation pada peubah respon.
English Abstract
Regression analysis is a method used to determine the relationship of dependency between the response variable (Y) with explanatory variables (X). Poisson regression model is an analysis for count data. Poisson regression is found on the assumption of an average value equal to the value of (equidispersion) but can be found data that has a value greater than the value of the mean (overdispersion). Poisson regression, so not exactly used to model data. The proper alternative models for those data are Generalized PoissonRegression (GP) and Negative Binomial Regression (NB). Response variables often encountered on any data is zero with large proportions (zero inflation), that is more than 50%. Large proportion of zero may result in accuracy (precision) of the inference. In addition, Poisson regression was also not appropriate to mention the actual data model. An alternative to the more appropriate regression model to data that experiencing overdispersion and has many zero values response variables are Zero Inflated Generalized Poisson Regression (ZIGP) and Zero Inflated Negative Binomial Regression (ZINB). The purpose of this research is comparing the performances of Zero Inflated Generalized Poisson Regression (ZIGP) and and Zero Inflated Negative Binomial Regression (ZINB) based on AIC values. The data used is 3 data secondary who suffered overdispersion and zero inflation on response variables. The result analysis shows a model Zero Inflated Generalized Poisson Regression (ZIGP) is more appropriate if used on data that experienced overdispersion and zero inflation on response variables.
| Item Type: | Thesis (Sarjana) |
|---|---|
| Identification Number: | SKR/MIPA/2013/382/051400161 |
| Subjects: | 500 Natural sciences and mathematics > 519 Probabilities and applied mathematics > 519.5 Statistical mathematics |
| Divisions: | Fakultas Matematika dan Ilmu Pengetahuan Alam > Statistika |
| Depositing User: | Hasbi |
| Date Deposited: | 17 Jan 2014 09:45 |
| Last Modified: | 25 Oct 2021 02:51 |
| URI: | http://repository.ub.ac.id/id/eprint/153649 |
Preview |
Text
ABSTRAK.pdf Download (1MB) | Preview |
Preview |
Text
BAB_1.pdf Download (1MB) | Preview |
Preview |
Text
BAB_3.pdf Download (1MB) | Preview |
Preview |
Text
BAB_4.pdf Download (2MB) | Preview |
Preview |
Text
BAB_2.pdf Download (2MB) | Preview |
Preview |
Text
BAB_5.pdf Download (1MB) | Preview |
Preview |
Text
COVER_DALAM.pdf Download (1MB) | Preview |
Preview |
Text
DAFTAR_ISI.pdf Download (1MB) | Preview |
Preview |
Text
DAFTAR_LAMPIRAN.pdf Download (1MB) | Preview |
Preview |
Text
DAFTAR_PUSTAKA.pdf Download (1MB) | Preview |
Preview |
Text
DAFTAR_TABEL.pdf Download (1MB) | Preview |
Preview |
Text
KATA_PENGANTAR.pdf Download (1MB) | Preview |
Preview |
Text
LEMBAR_PENGESAHAN.pdf Download (1MB) | Preview |
Preview |
Text
LEMBAR_PERNYATAAN.pdf Download (1MB) | Preview |
Actions (login required)
 |
View Item |