Open Fumire opened 2 years ago
Two suggestions
Objective: Balanced Accuracy > 0.8
Add other classification algorithms
How about using candidate taxa?
Add bagging algorithm
Separate by antibiotic usages
Improved by over-sampling
Over-sampling vs. Under-sampling
Classification starts with DATs
Add machine learning results