Predict the risk of death (mortality) in patients admitted to ICU within hospitals as part of the Xerox Mortality Prediction Challenge.
Details of the files are as follows:
- MortalityPredictioninICU-XRCI_MiniProject.pdf - details of the project
- featureExtract.py - extracts 79 features from data as described in MortalityPredictioninICU-XRCI_MiniProject.pdf
- neuralNetwork.py - 3 layer artifical neural network, with 30 neurons in both the hidden layers
- run_model.py - generating predictions of the model on test data
- score.m - scoring function calculating sensitivity score, specificity score and mean prediction time score