For your own notes. </ b> Use Random Forest to separate disease A and disease B. Find out which of the biomarkers you entered is important. In UCL's previous research, Accuracy was 74.0%, so the goal is to exceed that.
The code and description can be found at here. [Machine learning] Disease classification by Random Forest
This time, I tried to classify diseases using Random forest. In our experiment, it was 75.4%, 1.4% higher than the previous study's 74.0%. In the previous study, only Gray matter volume was used, but we input the map obtained from diffusion MRI. For the time being, I would like to investigate Sensitivity and Specificity.
Recommended Posts