标题：An Application of Genetic Algorithm Approach and Cobb-Douglas Model for Predicting the Gross Regional Domestic Product by Expenditure-Based in Indonesia
摘要：This paper empirically investigates the modelingof Gross Regional Domestic Product (GRDP) predictionusing a genetic algorithm approach and Cobb-Douglas modeland also studied the variables that influence it. The geneticalgorithm approach used to estimate the Cobb-Douglas modelparameters. Meanwhile, the Cobb-Douglas model utilized forpredicting the model of GRDP that measured using the valueof expenditure (expenditure-based). Further, we also discussedthe comparison of the level of prediction errors using thegenetic algorithm and ordinary least square (OLS) approaches.The results of the analysis show the level of prediction errorsin the sample assessed by using Mean Absolute PercentageError (MAPE) that obtained from the estimator using geneticalgorithms approach is 0.177, while by using the OLS approachis 0.190. It shows that the prediction value of errors from agenetic algorithms approach is smaller when compared to theOLS approach. Of these, we found that the genetic algorithmis the best approach used to estimate the model of GRDP thatmeasured using the value of expenditure in Garut Regency,Indonesia. Also, by using a genetic algorithm approach, thisstudy also found that the variables of government consumptionexpenditure, gross domestic fixed capital formation, andnet export, and determined inventory changes, the GRDPprediction for the next two periods is IDR 44,449,327 and IDR45,457,805. Therefore, the model of GRDP prediction usingexpenditure-based values can be used as a consideration inmaking the budget plan by the local government in GarutRegency, Indonesia.
