Type Alias ModelMetricsBinary

Binary metrics for categorical values with two options. Details: https://help.qlik.com/en-US/cloud-services/Subsystems/Hub/Content/Sense_Hub/AutoML/scoring-binary-classification.htm

type ModelMetricsBinary = {
    accuracy?: number;
    accuracyTest?: number;
    auc?: number;
    aucTest?: number;
    f1?: number;
    f1Test?: number;
    fallout?: number;
    falloutTest?: number;
    falseNegative?: number;
    falseNegativeTest?: number;
    falsePositive?: number;
    falsePositiveTest?: number;
    logLoss?: number;
    logLossTest?: number;
    mcc?: number;
    mccTest?: number;
    missRate?: number;
    missRateTest?: number;
    npv?: number;
    npvTest?: number;
    precision?: number;
    precisionTest?: number;
    recall?: number;
    recallTest?: number;
    specificity?: number;
    specificityTest?: number;
    threshold?: number;
    thresholdTest?: number;
    trueNegative?: number;
    trueNegativeTest?: number;
    truePositive?: number;
    truePositiveTest?: number;
}
Index

Properties

accuracy? accuracyTest? auc? aucTest? f1? f1Test? fallout? falloutTest? falseNegative? falseNegativeTest? falsePositive? falsePositiveTest? logLoss? logLossTest? mcc? mccTest? missRate? missRateTest? npv? npvTest? precision? precisionTest? recall? recallTest? specificity? specificityTest? threshold? thresholdTest? trueNegative? trueNegativeTest? truePositive? truePositiveTest?

Properties

accuracy?: number

Average measure of how often the model made correct predictions (training data)

accuracyTest?: number

Average measure of how often the model made correct predictions (test data)

auc?: number
aucTest?: number
f1?: number
f1Test?: number
fallout?: number
falloutTest?: number
falseNegative?: number
falseNegativeTest?: number
falsePositive?: number
falsePositiveTest?: number
logLoss?: number
logLossTest?: number
mcc?: number
mccTest?: number
missRate?: number
missRateTest?: number
npv?: number
npvTest?: number
precision?: number

Positive predictive value. Probability that the model was correct when it predicted something was true. (training data)

precisionTest?: number

Positive predictive value. Probability that the model was correct when it predicted something was true. (test data)

recall?: number
recallTest?: number
specificity?: number
specificityTest?: number
threshold?: number
thresholdTest?: number
trueNegative?: number
trueNegativeTest?: number
truePositive?: number
truePositiveTest?: number

Settings

Member Visibility

On This Page

Properties