Supervised learning
A learning method that uses inputs together with their corresponding target values or classes as training examples.
What does a supervised model learn from?
Supervised learning uses inputs together with their corresponding targets for training. These targets are called labels and may be classes or numerical values. The model is to learn relationships that allow it to predict appropriate results for new inputs. The targets must fit the intended task. The term supervised refers to these training targets; human approval during later operation is defined as a separate process step.
In a fictional service project, previous messages with reviewed assignments to the Repair or Billing teams are to serve as training examples. The team first establishes whether these assignments reflect the responsibility wanted today and were applied consistently. Incorrect labels can teach incorrect relationships. Further suitable messages that are not part of training are planned for assessment. This should also examine how the model handles unclear or previously unfamiliar requests. Having labels enables this learning method, but does not establish sufficient quality for deployment.
Sources
Related terms
- Machine learningModels learn patterns from data and apply them to new inputs.
- AI inferenceApplying an already trained model to inputs to obtain a result.
- LLM evaluationLLM evaluation tests model outputs against defined tasks and criteria. One good example does not establish overall quality.
- Forecasting modelA model that uses available information to estimate a future quantity for a defined period.
- Unsupervised learningA learning method that looks for patterns in data without predefined target answers for the particular training task.
- Re-rankingRe-ranking reassesses search candidates already retrieved. Relevance to the query and the limits of the initial selection both matter.
Your project with ASES
Describe your project to us. We will get back to you with an initial assessment.