Prediksi Harga Material Bangunan Dengan Autoregressive Integrated Moving Average (Arima) Pada CV. TJA

The problem most often faced by CV TJA is that the price from the RAB (Cost Budget Plan) calculation is not in accordance with market prices, so it is necessary to predict the price of building materials to help the company prepare the RAB. The aim of this research is to examine predictions of material prices measured using the ARIMA method. The data analysis method in this study used the MA parameter (1) with the ARIMA model (0,1,1) for dynamix cement material, the MA parameter (1) ARIMA model (0,2,1) for cast sand, the MA parameter (1 ) ARIMA model (0,3,1) for threaded iron 13 and AR parameter (2) ARIMA model (2,1,0) for usuk wood material. By using 3 error testing methods, namely Mean Absolute Percentage Error (MAPE), Mean Absolute Deviation (MAD) and Mean Squared Deviation (MSD). The results of this analysis show the lowest accuracy value, namely for dynamic cement material with a MAPE value of 5%, a MAD value of 1.986 and an MSD value is 63.584.667. The results of error tests using MAPE, MAD and MSD can be concluded that the ARIMA method is very accurate because the MAPE value is less than 10%.