Perbandingan Model Regresi Spline Adaptif Berganda dan Model Probit dalam Peramalan Resesi  

Oleh:

Aziz, Azwirda

Jenis: Article from Journal - ilmiah nasional - tidak terakreditasi DIKTI
Dalam koleksi: Widya: Majalah Ilmiah vol. 24 no. 261 (Jun. 2007), page 12-20.
Topik: Multivariate Adaptive Regression Splines; MARS; Peramalan Resesi; Probit Model

Ketersediaan

Perpustakaan Pusat (Semanggi) Nomor Panggil: MM47.25 Non-tandon: 1 (dapat dipinjam: 0) Tandon: tidak ada

Lihat Detail Induk

Isi artikel

In modeling of recessions in Indonesia with some its connections with economic factor, most of economic variables will follow nonlinear cycle and the form of its curve is not known because its form is not simple or does not follow the familiar function that have been already been. While many of studies that have been done by the analiysts employed probit model to estimate the probability of recession. Probit model is one of global parametric models that are only accurate if the specified model is reasonable approximation to the true underlying function. Consequently, if probit model is employed to estimate the probability of recession it may not adequately capture the underlying processes related to recession, Based on that consideration it is needed a method that can solve the nonlinear problem and not needed to determine its prior curve form that is multivariate adaptive regression splines (MARS) used for prediction. The objective of this study was to compare multivariate adaptive regression splines model with Azwirda Aziz STIE Swadaya-Jakarta probit model in recession for casting based on financial and real variables in Indonesia. This study used five predictor variables to forecast recessioan and the analysis is undertaken on monthly data. Results showed that for insample recession forecasting, MARS model is much better than probit model whereas for out-of sample recession forecasting, MARS model is nearly as well as probit model.

Opini Anda

Klik untuk menuliskan opini Anda tentang koleksi ini!

Kembali