Application of GA Optimizing Neural Network to Coal Sales Forecasts

Abstract

With the rapid development of the market, coal enterprises predict sales by subjective experience, which is far from accurate. In order to minimize decision-making errors, to avoid warehouse inventory shortages or backlog and to increase prediction accuracy of coal sales forecast, the study of forecast methods is particularly important. In the paper, improved BP algorithm is adopted based on some large coal enterprises' practical characters. Connection weights are optimized by generic algorithm. The forecast method is implemented in the paper. Theoretical analysis and experimental results show that neural network is feasible and effective for coal sales prediction, with a bright future. Genetic algorithm optimizing neural network increases speed calculation and reliability.

Optimization Algorithm in Supply Chain Management

“…Sales forecasting is a method involved in practices which involves decision-taking and it plays a crucial role in goods and services marketing. Genetic algorithm was applied in optimizing predictions of coal sales via neural network optimization by Zhang and Qinghe [86]and Gao et al [57]developed an integrated model of decision tree algorithm and genetic algorithm for the same. In [89], authors presented GA for forecast of car sales.…”