PERPUSTAKAAN

The land value of information is needed by the government, the community, and the private sector. During this time the information of available land value is yet away from the market value, as a result it is causing a loss of potential income to occur against state revenue. This study was conducted to find out the best way for scoring variable model in the field-based land valuation data processing with the result that obtains accurate and specific land values. Determination of scoring is performed by the Analysis Hierarchy Process (AHP) method to obtain outcome a multiplier factor to the initial scoring of the predetermined independent variables. Sample estimation results are matched with samples which is obtained in the field so that the best model is formed from the calculation of mass land valuation applications with ANNAVAL Artificial Neural Network programming language.
The result of manipulation of scoring with AHP is particularly effective in producing of multipliers for each independent variable. The multiplier factor by initial scoring of the existing of independent variables is estimated with the application of ANNAVAL as a result to produce an estimated land value. Calculation of the estimated land value is chosen based on the value (R) of the evaluation target of the model and with a certain hidden layer and learning rate to produce an extra rational value. It is readily apparent based on the results of the validation carried out by calculating the application of the measure of the accuracy of the coefficient correlation value of 0.940107, the RMSE value of 491195.6533, the MAD value of 372773.6, although the results of the RMSE error as a whole the sample data state that (R) the evaluation target coefficient of the high artificial neural network model has more small value. While the MAPE value shows that 17.32% with accurate forecasting criteria. Coefficient correlation number 0.940107 with correlation between estimated value and actual number is more accurate.

Keywords: Scoring System Manipulation, Soil Assessment with ANN