This app was made in collaboration with Georgios N. Kalfas
Our job was to build a decision tree and a random forest (using the patient’s profile as features and the drug types as classes) to find out which drug might be appropriate for a future patient with the same illness. The Shiny app created for the task, has been made with non-proficient users in mind. A patient's profile can be manipulated (i.e., the age of the subject), based on the selected features a different classification will appear (please refer to the step-by-step explanations included in the app). Please, bear in mind that the app was created for illustrating purposes and in this example, sex has no significance on the classification.
- The app can be also found here
The data is collected from a set of patients who suffered from the same illness during a treatment, which consists of two parts: 1) the patient’s profile, including Age, Sex, Blood Pressure (BP), Cholesterol, and Sodium-potassium Pump (Na_to_K). 2) the five drugs (A, B, C, X, Y) that patients responded to.