|
Comparison of artificial neural networks and decision tree method to identify factors influencing dust storm (Case Study: Yazd province)
|
Mohsen Yousefi *   , Mohamadreza Ekhtesasi |
|
|
|
Abstract: (12923 Views) |
| One of the major natural disasters causing damage annually is desert. Dry lands of the world including Iran are whipped by strong winds and dust storms form several times each year. In this study, using data from meteorological stations (thunder storms, wind magnitude (size, amount), wind duration, visibility, fastest wind speed, average wind speed, prevailing winds and dust storms) in 1953-2005 were used. In order to determine the most appropriate combination of neural network and input parameters (inputs) influencing the phenomenon of dust storms from Variable reduction factor analysis (maximum likelihood, principal component), principal component analysis, stepwise progressive and gamma test were used. Each of the above mentioned methods, with a different combination of neural network with the algorithm of Levenberg-Marquardt have been used. The results showed that the stepwise progressive R² = 0.87 and RMSE = 0.04 provides the most suitable combination for a neural network. Comparison of simulated dust storm phenomenon in seasons and months different of the year showed that simulates the phenomenon of dust storm in summer and spring seasons and months of April, May, June, July, August and September, the accuracy is higher. In comparing with algorithm of Levenberg-Marquardt and decision tree with algorithm CART, Neural networks with a correlation coefficient of 0.87 and the root mean square error of 0.04 of the decision tree method with a correlation coefficient of 0.86 and the root mean square error of 0.06 has more carefully in order to simulate dust storm. |
|
| Keywords: dust storm, neural network, decision tree, algorithm CART, Yazd Province |
|
|
Full-Text [PDF 353 kb]
(1697 Downloads)
|
Type of Study: Research |
Subject:
Special
Received: 2015/07/28 | Accepted: 2015/07/28 | Published: 2015/07/28
|
|
|
|
|
|
|
| Send email to the article author |
|
|
|
| Add your comments about this article |
|
|
|
