%A NIM.: 19106010049 Indriani Wahyu Nur Pratiwi
%O Pembimbing: Sri Utami Zuliana, S.Si., M.Sc., Ph.D.
%T PEMILIHAN MODEL REGRESI B-SPLINE TERBAIK MENGGUNAKAN METODE AKAIKE INFORMATION CRITERION (AIC) PADA DATA RUNTUN WAKTU (STUDI KASUS : NILAI TUKAR MATA UANG RUPIAH (IDR) TERHADAP DOLAR AMERIKA (USD) PERIODE JANUARI 2018 - JANUARI 2023)
%X Currency exchange rate data is time series data, which is a collection of
data taken at the equally interval in time sequentially. One of the popular time series
data modeling is Autoregressive Integrated Moving Average (ARIMA). ARIMA
requires the assumption of stationarity, the assumption of normality, and the assumption
of white noise that must be met as conditions to form the ARIMA model.
If one of the assumptions is violated, B-spline regression can be an alternative. The
purpose of this study is to construct the model of exchange rate of rupiah against
the US dollar which can be used as a prediction of future exchange rates using
B-spline regression. The optimal B-Spline regression model is obtained based on
selecting the optimal combination of order and number of knot points using Akaike
Information Criterion (AIC) optimization. Furthermore, an estimate of the model
parameters was obtained and a model feasibility test was carried out to see the performance
of the model using the Mean Absolute Percentage Error (MAPE) value.
The combination of order and the optimal number of knot points is obtained in order
3 (quadratic) and 2 knot points with an AIC value of 857.8322 and a MAPE
value of 0.0148376. Optimal models that obtained is: ˆy = 13526.08N−2,3(x) +
14177.5N−1,3(x) + 14145.24N0,3(x) + 15698.26N1,3(x) + 15156.4N2,3(x).
%K AIC, MAPE, Regresi B-Spline, Runtun Waktu
%D 2023
%I UIN SUNAN KALIJAGA YOGYAKARTA
%L digilib59650