Pengelompokan produksi padi nasional dengan pendekatan data mining konsep K-means

Abstract

One indicator of national food security is rice production, generated from the crop area in irrigation area. Influencing policy is regulation and changes in strategic environment , which are both synergized in the form of irrigation system. JICA Study - FIDP , 1993, indicating that he development of irrigated area will move to region Sumatra , Kalimantan , and Maluku - Papua , but based on record of data production on the last 20 years the regions remains low. This study aims to find solutions the expectations of increasing rice production by analyzing which potential areas have to increase in production. Data Mining approach concept K -means method used in clustering to analyze the province in national rice production , analysis based on the 20 years data record rice production from 1993 to 2012 consists of 33 provinces whole of Indonesia , with the object of observation on the average production to the increase production (slope) and forecast production for year 2013. Based on result of simulation optimization with the K - Means obtained that rice production development sequence consisting of six clusters. Highest national rice production is still dominated by Java and Bali (group 1) , meaning that the development in this region are optimally exploit the potential and more effort is needed to maintain existing area by preventing land use change, but due to limited space (potential area development only 62000 Ha), the development of irrigation areas is more rational prioritized in groups 2, namely region Sulawesi , NTB, and West Sumatera. Simulation optimization has accommodate history of production of the dimensions past , current and future production.