K-Nearest Neighbors
← Back to Classical ML Algorithms
Instance-based learning that classifies new points based on the majority vote of their k nearest neighbors. No training phase — all computation happens at prediction time.
Key Properties
- Instance-based (lazy) learning
- Distance metrics: Euclidean, Manhattan, Cosine
- Choice of k controls bias-variance tradeoff
Related
- Feature Engineering (feature scaling critical for KNN)
- Dimensionality Reduction (curse of dimensionality affects KNN)