skip to main content

PEMODELAN REGRESI SPLINE MENGGUNAKAN METODE PENALIZED SPLINE PADA DATA LONGITUDINAL (Studi Kasus: Harga Penutupan Saham LQ45 Sektor Keuangan dengan Kurs USD terhadap Rupiah Periode Januari 2011-Januari 2016)


Citation Format:
Abstract

Nonparametric regression is one type of regression analysis used when parametric regression assumptions are not fulfilled. Nonparametric regression is used when the curve does not form a specific pattern of connections. One of the approach by using nonparametric regression is spline regression with penalized spline method. Spline regression using penalized spline method was applied to three closing stock prices on the financial sector such as Bank BRI, BCA and Mandiri with the data of USD currency rate in rupiah. Closing price of stock data and the USD currency rate in rupiah were taken from January 2011 up to January 2016 for in sample data and from February 2016 up to December 2016 for out sample data. The data taken is called longitudinal data which is observing some subjects on specific period. Best spline regression model with penalized spline method is derived from the minimum value of GCV, the number of optimal knots and the optimal orde. Best spline regression model with penalized spline method for longitudinal data was obtained on the orde of 1, the 59 knots, the smoothing parameter with λ value of 1 and the GCV value of 889,797. The R2 value of in sample data was 99,292%, best model performance for in sample data. MAPE value of out sample data is  1,057%, the best accurate performance model.

Keyword: stock price, USD currency rate, longitudinal data, spline regression, penalized spline

Fulltext View|Download
Keywords: stock price, USD currency rate, longitudinal data, spline regression, penalized spline

Article Metrics:

Last update:

No citation recorded.

Last update:

No citation recorded.