Afidati, Asna Della Nur (2018) Perbandingan Model Regresi Ridge Dan Partial Least Square (Pls) Untuk Mengatasi Masalah Multikolinieritas Dalam Regresi Linier Berganda. Sarjana thesis, Universitas Brawijaya.
Abstract
Analisis regresi linier berganda merupakan analisis yang digunakan untuk mengetahui hubungan antar peubah respon dan peubah prediktor. Apabila terdapat korelasi antar peubah prediktor dikatakan terjadi multikolinieritas pada data. Metode yang dapat digunakan untuk menangani masalah multikolinieritas yaitu regresi ridge dan regresi Partial Least Square (PLS). Penelitian ini bertujuan untuk mengetahui metode yang terbaik antara regresi ridge dan regresi PLS dalam mengatasi masalah multikolinieritas. Data yang digunakan adalah data Angka Kematian Bayi (AKB) di Jawa Timur tahun 2016 dan data harga penutupan saham perusahaan keuangan tahun 2015. Untuk membandingkan kedua metode digunakan kriteria kebaikan model R2 Adjusted, Root Mean Square Error (RMSE), dan Akaike Information Criterion (AIC). Hasil penelitian didapatkan nilai R2 Adjusted regresi PLS lebih besar daripada regresi ridge serta nilai RMSE dan AIC regresi PLS lebih kecil daripada regresi ridge. Sehingga dari nilai ketiga kriteria kebaikan model dapat diketahui bahwa regresi PLS lebih baik daripada regresi ridge.
English Abstract
Multiple linear regression analysis is an analysis used to determine the relationship between the response variables and predictor variables. Multicollinearity is condition where there is correlation between predictors variables. The methods that can be used to handle multicollinearity problems are ridge regression and Partial Least Square (PLS) regression. This study aims to determine the better method between ridge regression and PLS regression in overcoming multicollinearity problems. The data used are data of Infant Mortality Rate (IMR) in East Java in 2016 and closing price data of financial company stock in 2015. To compare the two methods used the goodness of fit model R2 Adjusted, Root Mean Square Error (RMSE), and Akaike Information Criterion (AIC). R2 Adjusted of PLS regression is bigger than ridge regression. RMSE and AIC of PLS regression are smaller than ridge regression. The result of this research is PLS regression better than ridge regression.
Other obstract
-
| Item Type: | Thesis (Sarjana) |
|---|---|
| Identification Number: | SKR/MIPA/2018/249/051807008 |
| Uncontrolled Keywords: | Multikolinieritas, Regresi PLS, Regresi Ridge, Multicollinearity, PLS Regression, Ridge Regression |
| Subjects: | 500 Natural sciences and mathematics > 519 Probabilities and applied mathematics > 519.5 Statistical mathematics > 519.53 Descriptive statistics, multivariate analysis, analysis of variance and covariance > 519.536 Regression analysis |
| Divisions: | Fakultas Matematika dan Ilmu Pengetahuan Alam > Statistika |
| Depositing User: | Nur Cholis |
| Date Deposited: | 03 Jun 2020 23:01 |
| Last Modified: | 22 Oct 2021 07:26 |
| URI: | http://repository.ub.ac.id/id/eprint/168544 |
Preview |
Text
ASNA DELLA NUR AFIDATI.pdf Download (1MB) | Preview |
Actions (login required)
 |
View Item |