Prediction of peatlands forest fire occurrence using extreme learning machine approach

In the literature, various machine learning approach has been considered to model the forest fire occurrences, including the Neural Networks approach. In this paper we consider an application of extreme learning machine (elm) method of the neural network model to predict the peatlands forest fire occurrence. For comparison purpose, we also compare elm performance with the "standard"Backpropagation (bp) learning approach. We found that the elm approach can obtain the prediction results much faster than the backpropagation method. However, in general, the elm method will be less accurate than the backpropagation learning approach