Advertisement
click left half = train, click right half = classify
Each click adds a training point (left) or queries (right). Color = predicted class.
What you're seeing
k-NN: for each query, find k nearest training points; predict the majority class.
Simple but powerful baseline. No training phase — but every query searches all training points (O(n) per query without index). Vector databases solve this with ANN indexes.
★ KEY TAKEAWAY
k-NN: for each query, pick majority class of k nearest training points. Simple but powerful baseline.
▶ WHAT TO TRY
- Click left half to add training points.
- Click right half to classify queries.
- Larger k = smoother boundary; smaller k = more local.