Disease Classification in R using Decision Tree (Random Forest) and Multinomial Logistic Regression

Algorithms Used:

• Random Forest
• Multinomial Logistic Regression

In the above project, we have done the following:

• Classified 42 diseases using Random Forest
• Acquired Feature Importance table and performed Feature Selection by selecting features above 75 percentile importance
• Used "New_Training" and "New_Testing" datasets to perform Multinomial Logistic Regression
• EDA using a series of Visualizations showing :
  1. Missing Values
  2. Correlation Matrix
  3. Symptom frequencies ... and more!

Final Accuracy Values of the Algorithms Used:

• Random Forest : 97.619%
• Multinomial Logistic Regression : 92.8571%

From the above information : Random Forest Algorithm is preferred for the given problem statement.