KNN Cluster AI

KNN Cluster AI applies adaptive nearest-neighbor clustering to recent price and flow-derived features. Clusters highlight where the market is structurally similar to its own recent history — pockets of recurring behavior, not smoothed overlays.

• Distance-based structure adapts as drift and volatility change
• Useful for context filters, regime alignment, and research workflows

Delivered for TradingView under your license. Dense charts and deep KNN settings can be CPU-heavy in the Pine editor — lighten lookbacks if the editor feels sluggish.

Follow the post-purchase email: license validation and TradingView script access. Keep TradingView updated within the supported major version for your license tier.

• k — number of neighbors (higher = smoother, less local)
• Feature set — which inputs feed the distance metric (e.g. returns, range, volume-normalized features)
• Lookback — how many bars feed the neighbor search
• Decay — optional weighting toward more recent observations

Cluster identity and distance-to-centroid style readouts are contextual, not buy/sell labels. Pair with your own rules for entry, sizing, and invalidation.

Clustering is descriptive, not predictive. Sparse data and structural breaks can produce unstable neighborhoods. No tool replaces judgment or risk controls.