Using over 5,800 images of chest radiographs, I utilized machine learning and neural networks to predict when pneumonia is present. The best model was able to predict over 80% accuracy on the test data with a false negative rate that was less than 3%.
