q2
Term · Chemistry · MLC-T-CHM-009712
A cross-validation statistic for a QSAR model, calculated as one minus the ratio of the predictive residual sum of squares to the total sum of squares of the observed activities, with each compound predicted by a model refitted without it. It is an internal measure of robustness and does not by itself show how well the model predicts an external test set, which needs a separate held-out statistic. A q2 above about 0.5 is conventionally regarded as acceptable.
| Identifier | MLC-T-CHM-009712 |
|---|---|
| Field | Chemistry |
| Subject | Chemistry and Human Health |
| References | PAC, 2016, 88, 239. 'Glossary of terms used in computational drug design, part II (IUPAC Recommendations 2015)' on page 254 (https://doi.org/10.1515/pac-2012-1204) |
| See also | r2 |
Record as JSON
{
"id": "MLC-T-CHM-009712",
"term": "q2",
"field": "Chemistry",
"definition": "A cross-validation statistic for a QSAR model, calculated as one minus the ratio of the predictive residual sum of squares to the total sum of squares of the observed activities, with each compound predicted by a model refitted without it. It is an internal measure of robustness and does not by itself show how well the model predicts an external test set, which needs a separate held-out statistic. A q2 above about 0.5 is conventionally regarded as acceptable.",
"subject": "Chemistry and Human Health",
"see_also": [
"r2"
],
"references": [
"PAC, 2016, 88, 239. 'Glossary of terms used in computational drug design, part II (IUPAC Recommendations 2015)' on page 254 (https://doi.org/10.1515/pac-2012-1204)"
],
"url": "https://mlchart.com/terminology/chemistry/q2/"
}
Record 10,319 of 13,676 in Chemistry terminology (MLC-0109). Request the full dataset.