Accuracy

How it is used

The simplest model metric, but misleading when classes are imbalanced: if only 1% of transactions are fraud, a model that always says 'not fraud' is 99% accurate. That is why teams also report precision and recall.

Example sentences

1. The model's accuracy is 97%, but it misses most of the cases we care about.

2. Accuracy dropped from 88% to 81% when we tested on last month's data.

Common collocations

• model accuracy
• improve accuracy
• test accuracy
• high accuracy

Related terms

• precisionOf all the cases a model labels as positive, the share that really are positive; high precision means few false alarms.
• recallOf all the cases that really are positive, the share that a model correctly finds; high recall means the model misses few real cases.
• classificationA type of modelling task in which the model predicts which category something belongs to, such as spam or not spam, or which product a review is about.
• test setA part of the data that is held back and never used in training, so that it can show at the end how well a model works on data it has not seen.

This word in other fields

The same English word can have a different meaning in another field:

• accuracy · Software Engineering
• accuracy · Artificial Intelligence & Machine Learning
• accuracy · Robotics & Automation
• accuracy · Journalism & Media

More terms: Modelling and Machine Learning

• model
• feature
• training data
• overfitting
• regression
• cross-validation
• baseline
• feature engineering
• target variable
• clustering