Confusion Matrix
Definition
A confusion matrix counts predicted and actual classes and makes error types visible.
In binary classification, it contains true positives, false positives, true negatives, and false negatives. From this, among others, precision, recall, and accuracy are derived.
Simply explained
Like a table showing which cases were sorted correctly and which into the wrong category.
Why it matters in practice
Relevant for understandable error analysis instead of just an overall metric.
Example from business
A complaint classifier shows which categories are frequently confused with each other.
Limits and pitfalls
- Only consider accuracy; read class order incorrectly; treat costs of different errors equally.
Origin & history
The matrix is a fundamental tool of classification analysis and is widely used in statistics and ML.
Use in mid-sized companies
- Possible application in SMEs: Check error types per class and derive data or process measures from them.
Frequently asked questions
Is the confusion matrix itself a metric?
It is a count table from which many metrics are calculated.
Sources
- confusion_matrix · scikit-learn, confusion_matrix
Sources checked on 10.10.2026.
Editorially reviewed on: