Forecasting Determinant of Cement Demand in Indonesia with Artificial Neural NetworkReport as inadecuate

Forecasting Determinant of Cement Demand in Indonesia with Artificial Neural Network - Download this document for free, or read online. Document in PDF available to download.

Journal: Journal of Asian Scientific Research

Abstract: This paper was presented Artificial Neural Network ANN as one of the predicting methods to obtain more accurate predicted data. Several methods have been applied to this purpose but still not gave a better accuracy. The method could be used in linear and nonlinear characteristic of data. Back propagation neural network was implemented in this experiment to solve the predicting problem. This research proposed to predict some future points to get the advantages from it. The data was demonstrated in generate from determinant of cement demand in Indonesia region. The data was used 6 variables which influenced the demand factors. The contribution of this experiment was an exploring the accuracy of predicted and the future points value of data. The result of this experiment was: data A, B, C, D, E and F data had the range of MSE 6.23e-9 until 1.34e-7. The MSE for the actual data was 2.39e-6 and the predicted was 1.99e-6. The predicting points has resulted on months 91 until 97 were 0.5854, 0.8448, 0.510, 0.6462, 0.8528, 0.516 and 0.5074 respectively. The delta between predicted and actual data were 0.0066, 0.1418, 0.069, 0.1492, 0.2038, 0.355, and 0.1046. The result of this experiment could be used to predict the future months on the time frame with neural network and measured the MSE of its performance. This study is one of very few studies which have investigated where it was specified in generated data from some variables of cement demand, time series demand characteristics of data, used neural network method with its parameters and in Indonesia region. Explanation of ANN step by step.

Other Journals

Journal of Asian Scientific Research

Month: 07-2015 Issue: 7

Related documents