Results of machine learning classification - DigiMed1-study

This document contains the confusion matrices of machine learning classification conducted in the context of the DigiMed1 study. In the qualitative part of the study, a coding scheme was first developed based on human ratings of open-ended responses. This coding scheme was then validated using automated classification methods (Naive Bayes, Logistic Regression, and Random Forest). The confusion matrices presented here provide an overview of the classification performance. unknown unknown