Classification of intensive care patients by using data mining methods

This study has been prepared to solve the problems in the intensive care classification system used in Turkey. Patient care in intensive care, patient follow-up, disease control, and the cost of all these are difficult procedures. Facilitating the control of this system is one of the most important aims of this thesis. The study includes the application of data mining methods. It was studied with real data of patients treated in intensive care units. Classification and clustering processes were performed on these data by taking classification values as reference. It also includes an application for predicting patients' mortality. The biggest problem experienced during the thesis study was the incomplete or incorrect way of recording the data. Before the data was ready for application, it had to go through a lot of pre-processing. Much more successful results will be achieved with instant data to be obtained with bedside monitors.