Open AccessDKI Jakarta vegetable food commodity inflation modeling with tsclust approach using k-error method
Author(s) -
I Made Sumertajaya,
Muhammad Nur Aidi,
Emeylia Safitri,
Yeni Rahkmawati
Publication year - 2020
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1567/2/022072
Subject(s) - commodity , inflation (cosmology) , mean squared error , cluster analysis , value (mathematics) , econometrics , cluster (spacecraft) , economics , statistics , mathematics , error correction model , k means clustering , computer science , cointegration , physics , theoretical physics , programming language , market economy
Inflation is a critical indicator in showing economic symptoms about prices in a region. The Province of DKI Jakarta is the nation’s capital city which plays an important role in the national economy including national inflation. An interesting commodity group is foodstuffs because it contains commodities that play a significant role in inflation in DKI Jakarta. TSClust approach was able to reduce the model by first bundling the commodities together. K-Error is a clustering method that incorporates the error element related to data to be clustered into the clustering step. The purpose of this study is to forecast inflation in Jakarta’s agricultural food commodities with the K-Error. Clustering with the K-Error resulted in a lower pseudo F value as the number of clusters increases. Next, the selection of optimum number of clusters was based on a stable pseudo F value, such as the value k = 4 to k = 8. Based on the RMSE value, the optimum number of clusters selected is 5 clusters with unique models. Based on the RMSE value, it can be seen that forecasting at the individual level produced RMSE values that were not significantly different from RMSE level of the cluster for each commodity.