Table 4 Model quality parameters shown in the Flame GUI
From: Flame: an open source framework for model development, hosting, and usage in production environments
Endpoint type | Parameter | Definition |
|---|---|---|
Qualitative | Sensitivity (fitting and prediction) | \(\frac{TP}{{TP + FN}}\) |
Specificity (fitting and prediction) | \(\frac{TN}{{TN + FP}}\) | |
MCC (fitting and prediction) | \(\frac{{\left( {TP*TN} \right) - \left( {FP*FN} \right)}}{{\sqrt {\left( {TP + FP} \right)\left( {TP + FN} \right)\left( {TN + FP} \right)\left( {TN + FN} \right)} }}\) | |
Quantitative | SDEC (fitting) | \(\sqrt {\frac{{\sum \left( {Y_{exp} - Y_{pred} } \right)^{2} }}{n}}\) |
SDEP (prediction) | \(\sqrt {\frac{{\sum \left( {Y_{exp} - Y_{pred} } \right)^{2} }}{n}}\) | |
r2 (fitting) | \(\frac{{\sum \left( {Y_{exp} - Y_{pred} } \right)^{2} }}{{\sum \sqrt {\left( {Y_{mean} - Y_{pred} } \right)} }}\) | |
q2 (prediction) | \(\frac{{\sum \left( {Y_{exp} - Y_{pred} } \right)^{2} }}{{\sum \sqrt {\left( {Y_{mean} - Y_{pred} } \right)} }}\) | |
Conformal models | Conformal coverage | \(\frac{Samples\;inside\;confidence\;boundaries}{{Total\;number\;of\;samples}}\) |
Conformal accuracy | \(\frac{Samples\;predicted\;correctly}{{Total\;number\;of\;samples}}\) | |
Mean interval (only quantitative) | \(\frac{{\sum \left| {Y_{max} - Y_{min} } \right|}}{n}\) |