A Benchmark on Artificial Intelligence Techniques for Automatic Chronic Respiratory Diseases Risk Classification

A major public health problem is the chronically respiratory ill patients. To create a more preventive and anticipatory system for these patients we can use artificial intelligence techniques. This work tackle the problem of developing a model for automatic classification of patients with risk of having a respiratory crisis on the biggest paediatric Public Hospital in Santiago, Chile. We present a benchmark of different approaches to create a model. The models were developed with history of biomedical signals for 45 patients from 0 months to 15 years old. We are able to identify to approaches which have a remarkable performance.