Open AccessIdentification of Potential Landslide Disaster in East Java Using Neural Network Model (Case Study: District of Ponogoro)
Author(s) -
Alvina Khairun Nisa,
Mohammad Isa Irawan,
Danar Guruh Pratomo
Publication year - 2019
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1366/1/012095
Subject(s) - landslide , artificial neural network , backpropagation , identification (biology) , java , natural disaster , geography , population , disaster area , computer science , seismology , geology , artificial intelligence , meteorology , ecology , demography , sociology , biology , programming language
Indonesia is the largest archipelagic country in the world, which is geographically located in areas prone to natural disasters. One of the frequent occurrences of natural disasters is a landslide. East Java Province is one of the areas that has the potential for landslides. This is due to the topography of the most mountainous and rugged territory. Besides that, it also caused high levels of population density in the region of hills so that raises pressure on ecosystems. The tendency of the occurrence of landslides in an area can be connected with the equality of land characteristics and climate in other regions on a landslide in the past. To reduce the risk of disaster will be designed the software-based neural network for identification of potential avalanche areas. With potential landslide identification software, it can help to identify other locations that have similar physical and soil characteristics, so that the area can be suspected of being a potentially landslide area. The overall test results of this study are using Backpropagation artificial neural networks with 7 inputs, 15 hidden and 1 output. The training function used is Resilient backpropagation (RP) with an accuracy of testing data is 90.56% and MSE of 0.0944.