<![CDATA[Bali Regional Income Board is a regional organization tasked with determining the amount of local tax revenue target for the next fiscal year. Currently still done manually in accordance with existing upgrading trends from previous years. So it needs to be done in way that can be measured and accurate forecasting. In recent studies, it shows that forecasting by combining conventional and artificial intelligence (hybrid) methods results in better forecasting accuracy. By that reason, the writer tries to forecast the target of revenue from Motor Vehicle Tax (PKB and BBNKB), which contribute 70% to Bali Province income by combining ARIMA method and Genetic Algorithm. The data used consisted of five groups: yearly and new Vehicles that have linear data types, and as Reverse Names, Entrance Mutations and Output Mutations that have non-linear data types. Each data group consisted of PKB and BBNKB, where it’s monthly realization data from 2011 to 2016 used to be training data and realization data for 2017 as test data. The Combined forecasting mechanism is performed using ARIMA to forecast linear data and using Genetic Algorithms for non-linear data. As a benchmark for combined forecasting using ARIMA and Genetic Algorithms, forecasting using ARIMA and Genetic Algorithms independently is used for all data types (linear and nonlinear). Testing is done by comparing data of forecasting result with that 3 different methods for year 2017 with data realization year 2017. Then the error percentage is counted using MAPE. From the test results obtained for ARIMA MAPE value of 3.63, Genetic Algorithm 4.72 and combined ARIMA and Genetic Algorithm of 1.13. Thus, the result of forecasting with combination ARIMA and Genetic Algorithm have the best result and then used to forecasting target of PKB for 2018 and so on]]>
