Demo 03 of 11
Diagnostic model comparison
AUC values: thesis results. ROC curve: synthetic data for illustration
- What it shows
- The AUC of four classifiers for each disease, as a table and bar chart, and an ROC curve where a threshold slider shows the sensitivity and false positive tradeoff.
- How it maps to my work
- The AUC values are the reported results of my M.Tech thesis on an external test set. The ROC curve is a binormal curve drawn to match the selected AUC, not the thesis curve itself. The thesis learned its decision thresholds to minimize false negatives, which is the tradeoff the slider shows. See the research write-up.
- Data source
- AUC values: thesis results. ROC curve: synthetic data for illustration.