Papers by Thomas Gebregergis
ICU recourse is limited due to high cost, thus, the intensive care need to be provided to the pat... more ICU recourse is limited due to high cost, thus, the intensive care need to be provided to the patients who need to be took care of most. Developing a good approach to predict the in-hospital mortality can be quite helpful on evaluating the need for ICU, and making triage decisions. Physionet 2012 challenge collected data from patients during the first 48 hours ICU stay, and called for good methods for predicting the mortalities. In our project, we applied some machine learning methods to learn the information in the large ICU dataset. Features, e.g., mean, variance, median, skewness, kurtosis, were extracted, and non-significant features were removed with some method, e.g., t-test, forward/backward/stepwise selection. Logistic regression model, support vector machine, and neural network model were trained to make prediction. The logistic regression model gives the best performance, and can correctly classify 42% percent of the all the patients in dataset C.
Uploads
Papers by Thomas Gebregergis