Lazy Learner Classifier

Machine learning algorithms are often grouped into eager learners and lazy learners based on when they actually do the work of learning from data.

Eager vs Lazy Learners

Eager Learner:
- Builds a general model during training (e.g. Logistic Regression, Decision Tree)
- Training takes longer, prediction is fast

Lazy Learner:
- Stores the training data and defers computation until prediction time
- Training is instant, prediction is slower

Why KNN is a Lazy Learner

KNN doesn't build an explicit model during training - it simply memorizes the training data. All the actual computation (finding neighbors and voting) happens when a new prediction is requested.

Trade-offs

Lazy learners adapt easily to new data since there's no retraining step, but they can become slow at prediction time as the dataset grows large, since every prediction requires scanning the stored data.

Coming Up Next

Next, you'll learn how to tune KNN's key hyperparameter - the value of k.