TRCC: T-Regulated Cytokine Clustering with Path-Density Mutual-Reachability Merging

We present TRCC v1.1, a density-peak clustering algorithm engineered for large-scale efficiency (‭$10^3 $‬‭$\hphantom{}\le n $‬‭$\hphantom{}\le 10^6$‬). The algorithm combines a kNN-truncated Gaussian density estimator with a novel path-density mutual-reachability merge condition that repairs over-segmentation by examining the cytokine-signal ridge between centroids.Key Results: Performance: Achieved a measured runtime of 3.45 seconds for 100,000 points. Accuracy: Attained a mean Adjusted Rand Index of 0.639 across a 37-dataset suite, outperforming K-Means and DBSCAN. Scalability: Implementation in C++ (via pybind11 and nanoflann) yields a 30x speedup over pure-Python fallbacks. Application: Demonstrates effective anomaly isolation in CAN-bus cyber-telemetry contexts.