|
|
 |
| |
|
|
|
Prediction of Landslides Using Data Mining Techniques |
|
|
|
PP: 437-447 |
|
|
|
doi:10.18576/amis/190217
|
|
|
|
Author(s) |
|
|
|
Raed Alazaidah,
Mohammad Subhi Al-Batah,
Ghassan Samara,
Mo’ath Alluwaici,
Mohammed Abu Safaqah,
Mohammad Aljaidi,
Suleiman Ibrahim Mohammad,
Asokan Vasudevan,
Yujie Zhang,
|
|
|
|
Abstract |
|
|
| Landslide is a phenomenon that can happen suddenly or slowly over a long period of time. It is defined as a mass movement of different materials such as rubbles or stones. Landslides not only may cause huge loss of lives, properties, livestock but also have a bad impact on the environment. Many classification models have been proposed and utilized in aim of prediction of landslides. Hence, this paper aims to conduct an extensive evaluation of large number of classification models that adopt several learning strategies using a primary dataset that has been collected specifically for this purpose. Moreover, another main objective of this paper is to deter-mine the best feature selection method to use among four well-known methods. The evaluation phase considers several related evaluations. The results revealed that RandomForest, had the best performance classifiers, with 87.73% accuracy. Moreover, Correlation Attribute Evaluator method showed the best predictive performance.
|
|
|
|
|
|
|
|
