R과 JAVA를 연동한 수요예측 모듈 구현- K사 은행자동화기기 사례중심으로 -

In this paper, the ATM demand forecasting module in the K bank has been implemented linking JAVA and R programs. Commercial statistical software has an easy-to-use advantage, however, costs too much, and lacks freedom to analyse user-specific statistics. Furthermore, it is very restricted for the existing statistical software to be linked to the enterprise database. This study proposes a implementation scheme to program the ARIMA statistical model programmed in R, and to be linked to enterprise database with a JAVA-based interface. In this study, unit root test, model identification, model estimation, and model diagnostic checking are automatically to be done by the JAVA program module which is capable of dealing with the enterprise database. This study shows that the proposed module practically be linked to JAVA based enterprise system in forecasting the ATM demand in the K bank.