ANALISIS REGRESI NONPARAMETRIK KERNEL MENGGUNAKAN METODE JACKKNIFE SAMPEL TERHAPUS-1 DAN SAMPEL TERHAPUS-2 (Studi Kasus: Pemodelan Tingkat Inflasi Terhadap Nilai Tukar Rupiah di Indonesia Periode 2004-2016)

Agum Prafindhani Putri, Rukun Santoso, Sugito Sugito


Exchange rate is a conversion between currencies of a country to another country. Inflation can be defined as the rise of good and service’s level of price continually. The fluctuation of exchange rate is related to inflation, because inflation is the reflection of changes in the price level which happens in market and led to changes in level of money demand and supply. From the data distribution pattern which doesn’t show linearity relation, therefore the right modeling needs to be done using non-parametrical regression. Kernel Function which is used in non-parametrical component is Gaussian with optimal choice of bandwidth using the delete-1 Jackknife sample and the delete-2 Jackknife sample in Cross Validation (CV) method. This research using monthly data, 100 in sample data which taken from September 2014 until December 2012, while the number of out sample data used is 40 which taken from January 2013 until April 2014. Based on the analysis which had been done, the best kernel non-parametrical regression is the model using the delete-2 Jackknife sample because it produced the smallest Mean Absolute Percentage Error (MAPE) therefore it had better model accuracy evaluation.


Keyword : Exchange Value, Non-parametrical Regression, Kernel, Jackknife Method, Cross Validation (CV)


Exchange Value, Non-parametrical Regression, Kernel, Jackknife Method, Cross Validation (CV)

Full Text:



  • There are currently no refbacks.

Creative Commons License
Jurnal Gaussian by Departemen Statistika Undip is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.

Flag Counter