Prediction of Forest Fire Risk to Trigger IoTs Reconfiguration Action

Nowadays, the ever-increasing demand for IoT nodes for adapting to changing environment conditions and application requirements rapidly raising the need of reconfiguring already existing IoTs nodes. Forest fires is one of the main causes of environmental degradation and its detection and prediction is a challenge, a case study of predicting IoTs reconfiguration. Each prediction algorithm has its own advantages and disadvantages, which lead to different predictive results for each specific IoT application reconfiguration behavior. It is important to determine which set of metrics are effective for predicting. The objective of this work is to choose the most suitable prediction algorithms for detection of Forest Fire Risk for trigger IoTs reconfiguration actions. In this work, a comparative study between various prediction algorithms is carried out. The performance metric is based on the accuracy, precision, recall, and the training time. The experimental results show that Feedforward Neural Network is the most accurate that gives a good prediction accuracy.