| dc.contributor |
Fakultas Matematika dan Ilmu Pengetahuan Alam |
|
| dc.creator |
Maulina, Silvia |
|
| dc.creator |
Karyana, Yayat |
|
| dc.date |
2020-09-03 |
|
| dc.date.accessioned |
2021-03-15T03:44:41Z |
|
| dc.date.available |
2021-03-15T03:44:41Z |
|
| dc.identifier |
http://karyailmiah.unisba.ac.id/index.php/statistika/article/view/24622 |
|
| dc.identifier |
10.29313/.v6i2.24622 |
|
| dc.identifier.uri |
http://hdl.handle.net/123456789/28822 |
|
| dc.description |
Abstract. Multicollinearity can make the estimator unstable and the residual variance increases so that the confidence interval for β tends to be wider and makes the t count in the partial test insignificant even though the resulting is large. Least absolute shrinkage and selection operators can shrink the regression coefficients from insignificant variables to zero and even zero and can select variables so that variables that have high correlation with other variables are selected from the model and the resulting model becomes easier to interpret. To simplify the computation of least absolute shrinkage and selection operator the least angle regression algorithm is used. This algorithm is more efficient to use and is designed to produce linear models. Based on inflation data and the factors that influence it, it is obtained from the independent variable jumlah uang beredar , nilai tukar , harga minyak dunia , indeks harga ekspor , indeks harga impor , upah pekerja , and harga beras only variables of jumlah uang beredar and upah pekerja that significantly influence inflation with α = 5% in the estimation using MKT. In the estimation of least absolute shrinkage and selection operators using the least angle regression algorithm, the best model chosen based on the 10-fold cross validation is in the second stage, namely with a minimum value of 0.2092029 and minimum is 0.851385423.Keywords: Metode Least Absolute Shrinkage and Selection Operator Method, Least Angle Regression Algorithm, Multicollinearity, 10-fold cross validation.Abstrak. Multikolinearitas dapat membuat penaksir tidak stabil dan varians sisaan membesar sehingga selang kepercayaan bagi cenderung lebih lebar dan membuat pada uji parsial tidak signifikan meskipun yang dihasilkan besar. Least absolute shrinkage and selection operator dapat menyusutkan koefisien regresi dari variabel yang tidak signifikan menuju nol bahkan tepat nol serta dapat melakukan seleksi variabel sehingga variabel yang memiliki korelasi tinggi dengan variabel lain terseleksi dari model dan model yang dihasilkan menjadi lebih mudah diinterpretasikan. Untuk mempermudah komputasi least absolute shrinkage and selection operator digunakan algoritma least angle regression. Algoritma ini lebih efisien digunakan dan didesain untuk menghasilkan model linear. Berdasarkan data inflasi beserta faktor-faktor yang memengaruhinya diperoleh dari variabel independen jumlah uang beredar , nilai tukar , harga minyak dunia , indeks harga ekspor , indeks harga impor , upah pekerja , dan harga beras hanya variabel jumlah uang beredar dan upah pekerja yang signifikan memengaruhi inflasi dengan pada taksiran menggunakan MKT. Pada taksiran least absolute shrinkage and selection operator menggunakan algoritma least angle regresion didapatkan model terbaik yang dipilih berdasarkan validasi silang 10-fold berada pada tahap kedua yaitu dengan nilai minimum sebesar 0.2092029 dan minimum sebesar 0.851385423. Kata Kunci: Metode Least Absolute Shrinkage and Selection Operator, Algoritma Least Angle Regression, Multikolinearitas, 10-fold cross validation. |
|
| dc.language |
id |
|
| dc.publisher |
Universitas islam Bandung |
|
| dc.rights |
Copyright (c) 2020 Prosiding Statistika |
|
| dc.source |
Prosiding Statistika; Vol 6, No 2, Prosiding Statistika (Agustus, 2020); 304-311 |
|
| dc.source |
Prosiding Statistika; Vol 6, No 2, Prosiding Statistika (Agustus, 2020); 304-311 |
|
| dc.source |
2460-6456 |
|
| dc.source |
10.29313/.v6i2 |
|
| dc.subject |
Statistika |
|
| dc.subject |
Metode Least Absolute Shrinkage and Selection Operator, Algoritma Least Angle Regression, Multikolinearitas, 10-fold cross validation. |
|
| dc.title |
Penerapan Least Absolute Shrinkage and Selection Operator Menggunakan Algoritma Least Angle Regression dalam Penanganan Multikolinearitas pada Data Inflasi |
|
| dc.type |
info:eu-repo/semantics/article |
|
| dc.type |
info:eu-repo/semantics/publishedVersion |
|
| dc.type |
Peer-reviewed Article |
|
| dc.type |
kuantitatif |
|