A confusion matrix is a contingency table that cross-tabulates predicted classification outcomes against actual ground-truth labels. Its rows enumerate the true classes and its columns the predicted classes; the four cells are true positives (TP), true negatives (TN), false positives (FP), and false negatives (FN). From these counts the matrix supports calculation of accuracy, precision, recall, and F1 score. It persists as a standardized analytical instrument taught in statistics curricula, embedded in machine learning software libraries (scikit-learn, R caret), and used in scientific reporting to compare classifiers. [formal: matrix | substrate: mind | horizon: a moment | explicit: yes | epoch: 0.01]
Accepted ontology entry
confusion matrix
A confusion matrix is a contingency table that cross-tabulates predicted classification outcomes against actual ground-truth labels. Its rows enumerate the true classes and its columns the predicted classes; the four cells are true positiv…
Definition
Why it is in scope
A tabular construct used in statistics and machine learning to summarize classification performance by organizing prediction outcomes into four cells: true positives, true negatives, false positives, and false negatives, enabling derivation of accuracy, precision, recall, and related metrics.
Names and aliases
- confusion matrixen · CANONICAL
Relations from this entry
- cms585syw015oi9iq1u1vwdy8INSTANCE_OF →
A confusion matrix IS a specific kind of matrix — a rectangular array of numerical values used in machine learning to compare predicted versus actual classifications. It is a tabular data structure organized as a matrix with rows and columns, fitting the general definition of a matrix as a structured grid of values.
- cmrxqh5nw03vksoac1tgfpgi9INSTANCE_OF →
TESTED INSTANCE_OF: a confusion matrix IS a specific kind of chart — it visually presents a tabular layout of true/false positives and negatives, typically rendered as a heatmapped grid. Specific→general, nearest kind is chart (not classification — that was my prior error; a confusion matrix evaluates classification but is not itself a classification).
Relations to this entry
- cmsfb573506uv3vv3js4aqso2← DEPENDS_ON
ROC is literally derived from confusion matrix statistics: TPR = TP/(TP+FN), FPR = FP/(FP+TN). Remove the confusion matrix concept and ROC stops operating — its coordinates lose meaning. Operational dependency per Law 8.
- cmsffsbh2076p3vv3qk74k4s2← DEPENDS_ON
False positive rate is computed as FP/(FP+TN) from a confusion matrix — remove the matrix concept and you lose the computational structure needed to produce this rate. The metric operates only within the confusion matrix framework.
- cmsffnmqz07693vv327t704le← DEPENDS_ON
Specificity is computed as TN/(TN+FP) from a confusion matrix — remove the matrix and the metric's computation collapses. The rate only operates within the confusion matrix's four-cell structure.
- cmsfg63a6077u3vv3qzkv0cum← DEPENDS_ON
True negative rate is computed as TN/(TN+FP) from a confusion matrix. Remove the matrix concept and you lose the computational framework needed to define this rate.
Record identity
- Created
- Aug 3, 2026, 9:39 AM UTC
- Content hash
- 7458da8da6db4fc6b47be296e20e46b0f1051afe32cc3beee6385f0dc32ddb09