Abstract:
Value Added Tax revenue is one of resources for governmentbudgeting. The effectiveness of forecasting VAT revenue willreflect the government budgeting. Data Mining techniqueshas been applied in several application areas, Includedresearch forecasting VAT. However, factors. or variablesused in previous study do not consider other variables.This research focuses on other factors that may affect theforecasting. Various models have been built on many conditions.The study found that the Multilayer perceptron model whichused the variable value to the consumer price index, theimport products index, the private consumption index, theprivate investment index and the public expenditure wouldget value the VAT collect estimations, that were the mistakevalue from the estimations at the minimum average and 98 percent for the correct average.