AI engineering glossary · ML fundamentals
What is weak supervision?
Weak supervision (also called Distant supervision): Building training labels cheaply from imperfect sources such as rules, heuristics, existing databases or other models rather than from careful manual labelling. The labels are noisy, but at scale they can train useful models, often after their errors are modelled and combined.
Weak supervision: an example
Tagging support tickets as 'refund' whenever they contain the word 'refund' or 'money back' produces rough labels for thousands of tickets in minutes.
Related terms
More ML fundamentals terms
- Ablation study
- Active learning
- Akaike information criterion
- Annotation
- Baseline model
- Bayesian information criterion
- Bias error
- Bias-variance tradeoff
Where you learn it
Weak supervision is part of the AI Engineering Program, which teaches it with notes, a hands-on lab and an exam. The first lesson is free.