Aplicación de técnicas de aprendizaje automático para predecir áreas en riesgo de incendio forestal

Abstract

Abstract: Wildfires cause many environmental disasters and endanger people's lives. This Degree Project seeks the incorporation of machine learning techniques in worlds of artificial intelligence (algorithms), with the aim of involving humanity, which, due to lack of knowledge, do not have the possibility of developing predictive models that provide a solution to natural disasters such as forest fires. For this, an investigation of various techniques, methods, models (linear regression, neural networks) and strategies that Machine Learning uses was carried out, to this is added an investigation on dataset that reflect data from forest fires in this case data was used from the United States to find out the factors that cause it and to come up with a prediction of in which area a fire could be taking place in order to ensure the success of the Degree Work.As a result, a prediction model will be obtained, which shows the real data and the predicted data where the fire could occur and in turn the area of this.