Ανάλυση ιατρικών δεδομένων
View/ Open
Keywords
Μηχανική μάθηση ; Ανάλυση δεδομένων ; Feature selection ; Ανάλυση επιβίωσης ; Causal analysis ; Αλγόριθμοι ; Random forest ; Decision tree ; KNN ; Survival analysis ; OLS ; Matching ; Weighting ; Ιατρικά δεδομέναAbstract
The current Thesis is dealing with the study and analysis of data of patients suffering from Chronic Kidney Disease. Using Descriptive Statistics, aspects of the disease are investigated, such as the most common cause of death, the prescribed drugs, and at what ages there is higher mortality, etc.
Machine learning techniques are used to predict patients' life expectancy, cause of death, and the financial burden of the disease on patients, using features that were refined through Feature Selection techniques.
Survival analysis is performed to determine the life expectancy of patients depending on the stage they are in, either on dialysis or transplant. Finally, using Causal Analysis, it was investigated whether the drug ‘Trombyl’, depending on its dosage, prolongs the life of patients.