Matthew's Correlation Coefficient
GPTKB entity
Statements (60)
| Predicate | Object |
|---|---|
| gptkbp:instance_of |
gptkb:analysis
|
| gptkbp:evaluates |
evaluating predictions
evaluating risk |
| gptkbp:function |
four values
|
| https://www.w3.org/2000/01/rdf-schema#label |
Matthew's Correlation Coefficient
|
| gptkbp:introduced_in |
gptkb:G._P._Matthews
|
| gptkbp:is_a_key_metric_in |
gptkb:Artificial_Intelligence
|
| gptkbp:is_a_preferred_metric_for |
binary outcomes
|
| gptkbp:is_a_standard_measure_in |
machine learning research
|
| gptkbp:is_a_standard_metric_in |
data science
|
| gptkbp:is_a_tool_for |
model evaluation
|
| gptkbp:is_analyzed_in |
classification models
the performance of classifiers in imbalanced datasets the accuracy of predictions the performance of algorithms the performance of classifiers the reliability of predictions |
| gptkbp:is_applicable_to |
confusion matrices
|
| gptkbp:is_calculated_as |
(TP * TN -FP * FN) / sqrt((TP + FP)(TP + FN)(TN + FP)(TN + FN))
|
| gptkbp:is_compared_to |
different models
the effectiveness of different algorithms |
| gptkbp:is_essential_for |
imbalanced datasets
|
| gptkbp:is_evaluated_by |
false positives
binary outcomes false negatives true positives confusion matrix values true negatives |
| gptkbp:is_influenced_by |
sample size
|
| gptkbp:is_recommended_by |
accuracy in some cases
|
| gptkbp:is_related_to |
gptkb:Cohen's_kappa
|
| gptkbp:is_used_in |
gptkb:sports_team
gptkb:machine_learning research studies clinical trials environmental studies financial modeling predictive analytics quality control bioinformatics data mining medical diagnostics social sciences image classification text classification |
| gptkbp:key_concept |
statistics
|
| gptkbp:key_feature |
model effectiveness
|
| gptkbp:known_as |
gptkb:MCC
|
| gptkbp:measures |
predictive accuracy
the strength of association the correlation between observed and predicted values the agreement between two raters the predictive power of a model the quality of binary classifications |
| gptkbp:range |
-1 to 1
|
| gptkbp:sensitivity |
class imbalance
|
| gptkbp:type_of |
correlation coefficient
|
| gptkbp:used_in |
binary classification
|
| gptkbp:bfsParent |
gptkb:GLUE_Benchmark
|
| gptkbp:bfsLayer |
6
|