false negative
In plain wordsA miss: a test or AI system gives the all-clear, for example no illness, no fraud or no danger, when the problem is in fact there.
A case that a classifier or test labels negative although its true label is positive: a cancer missed on a scan, a fraudulent payment approved, a harmful post left up. In hypothesis testing it corresponds to a type II error, failing to reject a null hypothesis that is false. False negatives are counted in the confusion matrixconfusion matrix, n.: A table that counts a model's right and wrong answers by type, showing for each true category how often each category was predicted, so you can see exactly which things it mixes up.; the false negative rate, FN/(FN + TP), also called the miss rate, equals 1 − recallrecall, n.: Of all the items that really were positive, the share the model managed to find; a cancer screen with high recall misses very few people who actually have cancer.. They are often the costlier error in screening, safety and security, where a miss can be irreversible, so such systems are tuned for high recall and accept more false positivesfalse positive, n.: A false alarm: a test or AI system raises a flag, for example for illness, fraud or danger, when nothing of the kind is really there., typically sending flagged cases to human review. Because false negatives are by definition not flagged, they are easy to overlook after deployment: measuring them requires audits of cases the system let through, follow-up of outcomes or deliberately seeded test cases. Lowering the decision threshold reduces false negatives at the price of more false alarms.
An audit of 5,000 approved transactions found 12 false negatives that the fraud model had waved through.
Origin
Like 'false positive', the term arose in medical and laboratory testing in the first half of the 20th century; its statistical counterpart, the error of the second kind (type II error), was formalized by Jerzy Neyman and Egon Pearson between 1928 and 1933.
Related
- false positivefalse positive, n.: A false alarm: a test or AI system raises a flag, for example for illness, fraud or danger, when nothing of the kind is really there.
- recallrecall, n.: Of all the items that really were positive, the share the model managed to find; a cancer screen with high recall misses very few people who actually have cancer.
- confusion matrixconfusion matrix, n.: A table that counts a model's right and wrong answers by type, showing for each true category how often each category was predicted, so you can see exactly which things it mixes up.
- ROC curveROC curve, n.: A graph showing how a detector's hit rate and false-alarm rate change together as it is made more or less trigger-happy; the closer the curve hugs the top-left corner, the better.
- specificityspecificity, n.: Of all the cases that really were negative, the share a test correctly cleared; a highly specific test rarely raises false alarms on healthy people.
- human-in-the-loophuman-in-the-loop, adj.: Describes AI systems in which a person checks, corrects or approves the machine's work at key points, rather than letting it act entirely on its own.
- false positivefalse positive, n.: A false alarm: a test or AI system raises a flag, for example for illness, fraud or danger, when nothing of the kind is really there.
Sources
- Neyman, J. & Pearson, E. S. (1933). On the Problem of the Most Efficient Tests of Statistical Hypotheses. Philosophical Transactions of the Royal Society A 231.
Mentioned in
false positivefalse positive, n.: A false alarm: a test or AI system raises a flag, for example for illness, fraud or danger, when nothing of the kind is really there., confusion matrixconfusion matrix, n.: A table that counts a model's right and wrong answers by type, showing for each true category how often each category was predicted, so you can see exactly which things it mixes up., recallrecall, n.: Of all the items that really were positive, the share the model managed to find; a cancer screen with high recall misses very few people who actually have cancer.