A Model for Pest Infestation Prediction in Crops Based on Local Meteorological Monitoring Stations

The use of pesticides for controlling the population growth of a given pest is the most commonly used technique. One of the reasons for this is the response time that such method has, acting quickly and eliminating the threat to the crops. However, pesticides are known for their risk in the health of consumers, farmers, and pesticide appliers. Infestations are caused by insects, and many of these insects have characteristics that are strongly influenced by weather factors, for example, being ectothermic, which makes them fragile to temperature changes in the region. Based on this, it is possible to state that the behavior of insects is predictable, capable of being determined by the climatic changes in the region. In this context, this work proposes the creation of a model for prediction of an infestation based on the climatic changes in the region for three different insect species: Anastrepha fraterculus, Ceratitis capitata, Grapholita molesta, using information such as temperature, humidity, rainfall, and wind speed. To accomplish this work, data mining techniques were used for data preprocessing and machine learning algorithms were used to create the prediction model. In order to predict infestations, we make comparisons between methods such as decision trees, Bayesian learning, support vector machines and neural networks. The results obtained by the experiments done showed that the neural networks models achieved the best areas under the ROC Curve for all three species: 0.92, 0.96, and 0.97, while the decision tree model presented most of the worst results, with areas of 0.72, 0.83, and 0.71. At the end, the neural networks models made it feasible to predict possible future infestation with 4 days of anticipation.