Skip to content
    Flow-alerta.cab

    Accuracy

    AI fundamentals & modelsHit rate
    Editorially reviewed on: 2 sources
    Your reading view

    The short answer · 11 words

    In short

    Accuracy is the proportion of correct predictions in all evaluated cases.

    Flow-alerta.cab (2026): Accuracy. https://flow-alerta.cab/en/glossar/accuracy. Editorially reviewed on: 10 October 2026.

    How much detail would you like?

    Practical

    Easy to understand, but without class distribution and error costs often insufficient.

    Hypothetical calculation example: out of 100 cases, 95 are uncritical and 5 are critical. If a model always says “uncritical”, it reaches 95% Accuracy, but does not detect a critical case.

    Evaluate Accuracy along with class distribution, baseline, and the cost of various errors.

    At a glance

    Core
    In single-label classification, Accuracy is the number of correct predictions divided by all scored cases. Binary applies: (TP + TN) / (TP + TN + FP + FN). In rare classes, a high degree of accuracy can occur despite poor detection.
    Use
    Easy to understand, but without class distribution and error costs often insufficient.
    Example
    Hypothetical calculation example: out of 100 cases, 95 are uncritical and 5 are critical. If a model always says “uncritical”, it reaches 95% Accuracy, but does not detect a critical case.
    Limits
    Ignore class imbalance; optimize test data multiple times; conceal confidence intervals and subgroups.

    In context

    Classification Error Rate

    The classification error rate is the proportion of incorrect class decisions in all evaluated cases.

    F1-Score

    The F1 score is the harmonious means of precision and recall.

    Questions about this term

    Is high accuracy always meaningful?

    No, especially for unequal classes and error costs.

    What is the error rate?

    For the same single-label evaluation, 1 − Accuracy applies.

    All details and background

    In single-label classification, Accuracy is the number of correct predictions divided by all scored cases. Binary applies: (TP + TN) / (TP + TN + FP + FN). In rare classes, a high degree of accuracy can occur despite poor detection.

    ELI5 – Explained simply

    It shows how often the model is correct overall.

    Why it matters in practice

    Easy to understand, but without class distribution and error costs often insufficient.

    Example

    Hypothetical calculation example: out of 100 cases, 95 are uncritical and 5 are critical. If a model always says “uncritical”, it reaches 95% Accuracy, but does not detect a critical case.

    Pitfalls and limits

    • Ignore class imbalance; optimize test data multiple times; conceal confidence intervals and subgroups.

    Origin & history

    Accuracy is a basic measure of statistics and classification.

    Use in mid-sized companies

    • Evaluate Accuracy along with class distribution, baseline, and the cost of various errors.

    Related terms

    • Error rate
    • Precision
    • Recall
    • F1 score
    • Confusion matrix

    Matching Flow-alerta service

    AI opportunity analysis & roadmapAI consulting: Examine tasks, data and prerequisites of a possible AI use case.

    Matching tool / use case

    View use casesscikit-learn accuracy_score; confusion matrix, precision and recall.

    Sources

    Quellenprüfung: 10.10.2026. Definition und dokumentierte Einordnung abgeglichen; Beispiele beschreiben mögliche Anwendungen.

    Editorially reviewed on:

    Next step

    Put knowledge into practice

    Discuss your projectView use cases