Glossary term
Glossary term
Evaluation and Benchmarks
A value indicating how far apart the average of predictions is from the average of labels in the dataset.
Not to be confused with the bias term in machine learning models or with bias in ethics and fairness.
Created for this library
A churn modeling team monitors prediction bias as the difference between predicted average churn and observed average churn each month.
A pricing analytics team monitors prediction bias against historical demand to detect systematic over- or under-prediction.
An ad-tech team monitors prediction bias on calibrated click probabilities to detect drift between the model and the bidder.
Definition source: Google for Developers Machine Learning Glossary | Creative Commons Attribution 4.0 License