Winda Try Astuti, ILKOM UNNES and Much Aziz Muslim, ILKOM UNNES and Endang Sugiharti, ILKOM UNNES (2019) The Implementation of The Neuro Fuzzy Method Using Information Gain for Improving Accuracy in Determination of Landslide Prone Areas. Scientific Journal of Informatics , 6 (1). pp. 95-105. ISSN 2407-7658

Preview	PDF Download (3MB) | Preview
Preview	PDF (The Implementation of The Neuro Fuzzy Method Using Information Gain for Improving Accuracy in Determination of Landslide Prone Areas ) - Published Version Download (2MB) | Preview

Abstract

The accuracy of information is increasing rapidly as technological development. For the example, the information in determination of disaster severity. The disasters that can be determined is landslide. This determination can be conducted using the fuzzy method. One of method is neuro fuzzy. Neuro fuzzy is a combined method of two systems, fuzzy logic and artificial neural network. The accuracy of neuro fuzzy method can be increased by applying the information gain. The purpose of this study is to implement and to know the accuracy of the implementation of information gain as the selection of landslide data features. It conducted to the neuro fuzzy method in determining landslide prone areas. The distribution of training data and testing data was using 20 k-fold cross validation. The implementation of the neuro fuzzy method on landslide data was obtained an accuracy of 81.9231%. In the implementation of the neuro fuzzy method with information gain was conducted in classification process. The process will stop when the accuracy has decreased. The highest accuracy result was obtained of 88.489% by removing an attribute. So, it can be concluded the accuracy increase of 6.5659% in the implementation of the neuro fuzzy method and information gain in determination of landslide prone areas.

Actions (login required)

View Item