Prediction of national strategic commodities production based on multi-Response nonparametric regression with fourier series estimator

The Ministry of Agriculture in Indonesia stated that there are 11 strategic commodities that have the largest contribution to food security and the formation of inflation rates in Indonesia. There are rice, corn, shallot, garlic, red chili, cayenne pepper, beef, chicken, broiler eggs, sugar, and cooking oil. The supply of strategic commodities that are suitable to the needs of the Indonesian people can maintain the stability of national food security. Indonesia's Government depends on provinces that become the main producer of most commodities, like East Java. However, a prediction can be made to determine the availability of these commodities in the coming period, based on data from the previous period. Because the data has an oscillation pattern, Fourier series estimators in multi-response case is used to forecast. Fourier series have the flexibility to approach the data pattern smoothly. The data from the East Java Province Government in Indonesia is taken for 11 commodities. The result is an optimal model based on the parsimony model with the small Mean Square Error (MSE), a Generalised Cross Validation (GCV) and the big determination coefficient value. The model that has been selected has a small goodness of fit criteria to forecast. So, Fourier series estimators with a multi-response case is suitable to predict national strategic commodity production in East Java, that give high contributions to Indonesia's achievement for food security.