🛡️ One-Class SVM Anomaly Boundary: RBF Gamma and Nu (2D Contour Map)
2D contour map of a one-class SVM (RBF kernel) solved by SMO: decision-value heat map, f = 0 boundary, support vectors, flagged points, distribution of f and sorted dual weights, with six numerical checks.
⚙ Under the hood
2D contour map of a one-class SVM (RBF kernel) solved by SMO: decision-value heat map, f = 0 boundary, support vectors, flagged points, distribution of f and sorted dual weights, with six numerical checks.
anomaly-detectionone-class-svmrbf-kernelmachine-learningnovelty-detection
2D · HTML5 Canvas 2D · 60 FPS target · runs fully client-side, no install