ANN approach for predicting economic trends based on electric energy consumption during natural disaster period

Economy trend (eco-trend) is the most important factor for developing the country. Unfortunately, various inevitable and unpredictable factor causes an effect on economic trend while the Natural Disaster period happened. The fluctuation of the trend is then occurred and make it more difficult to forecast. According to this research, the analysing method of the eco-trends prediction was represented by stock prices prediction and use the datasets of some industrials sector which mainly uses electricity for production. Then we found that the stock prices can be predicted more precisely after increasing electric energy consumption to be input features taking by using Artificial Neural Network. However, the result of the prediction is precisely in the normal period only. Therefore to analyse the prediction occurring in natural disaster period (the flood of Thailand 2011), the crosschecking method is considered. Finally, For the performance comparison of experiment results, the least mean squares (LMS) and root mean squared error (RMSE) are used. Finally the results of this research, they not only show how power consumption makes the results of stock prediction are more precise, but also provide the time-delay that is the indicator of the economic trends changing and can then explain the behaviour of the industrial segment in the natural disaster period.