relation: https://digilib.uin-suka.ac.id/id/eprint/59018/
title: PERBANDINGAN OPTIMISASI MODEL REGRESI B-SPLINE MENGGUNAKAN METODE AKAIKE INFORMATION CRITERION(AIC) DAN METODE GENERALIZED CROSS VALIDATION(GCV) (STUDI KASUS : INFLASI KOTA YOGYAKARTA PERIODE JANUARI 2018 - JANUARI 2023)
creator: Ibnu Fadilah, NIM.: 19106010020
subject: Matematika
description: Predicting future inflation can be done by creating a model based on historical  inflation data. Inflation data is one of the time series data that usually has  a certain model. One way to model time series data is to use nonparametric regression.  One of the estimation techniques in nonparametric regression is B-Spline  estimation. B-Spline method has a function, with the basis of functions that are  commonly used, one of the reasons for using the B-Spline regression is because it  has good flexibility and the degree of the polynomial does not depend on the control  point but depends on the order of the polynomial, which means if you change the  data at one point the knots won’t change all the curves. After getting the combination  of order and knot points, the researcher optimizes to get the optimal order and  knot point. There are many methods that can be used to get the optimal combination.  In this study, researchers used the Akaike Information Criterion (AIC) and Generalize  Cross Validation (GCV). After getting the optimal order and knot point, we will  get an estimation model for the B-Spline. To get the optimal model, the researcher  uses the Mean Absolute Prediction Error (MAPE), which has a minimum value.  By using D.I Yogyakarta year-on-year inflation data from January 2018 – January  2023, the optimal models in this study are on the order of 4 (cubic) with 4 knots,  with AIC value 21,53983 and MAPE value 54,48514 optimal model that obtained  is: ˆ Y = (0.1810162)N−3,4(x) + (0.08516874)N−2,4(x) + (0.455261)N−1,4(x) +  (0.455261)N0,4(x)
date: 2023-03-31
type: Thesis
type: NonPeerReviewed
format: text
language: id
identifier: https://digilib.uin-suka.ac.id/id/eprint/59018/1/19106010020_BAB-I_IV-atau-V_DAFTAR-PUSTAKA.pdf
format: text
language: id
identifier: https://digilib.uin-suka.ac.id/id/eprint/59018/2/19106010020_BAB-II_sampai_SEBELUM-BAB-TERAKHIR.pdf
identifier:   Ibnu Fadilah, NIM.: 19106010020  (2023) PERBANDINGAN OPTIMISASI MODEL REGRESI B-SPLINE MENGGUNAKAN METODE AKAIKE INFORMATION CRITERION(AIC) DAN METODE GENERALIZED CROSS VALIDATION(GCV) (STUDI KASUS : INFLASI KOTA YOGYAKARTA PERIODE JANUARI 2018 - JANUARI 2023).  Skripsi thesis, UIN SUNAN KALIJAGA YOGYAKARTA.