Internal Department Budget Forecasting of a Company: Benchmark of Artificial Neural Networks

Budget forecasting plays an essential role in management activities. Governments, corporations, and companies want to know what they are spending in the future periods. It helps plan the investments, produces, human resources and affects companies’ future. The purpose of this paper is to estimate the budget for a company. Independent variables of the study are the last month and the last 12 months of the budget. Different training errors, hidden layers, dividing data are examined to determine the proper model. The neural network is used for forecasting, and results are shown with mean absolute percent error. The results showed that the number of hidden layers and dividing the data as a training-validation-test decreased the error in the budget forecasting.