Gaussian Mixture Model: Expectation-Maximization Clustering

This simulator fits a Gaussian Mixture Model to a 2D point cloud with real Expectation-Maximization, rendered as a tilted 3D scene you can orbit around. Each of the K components is drawn as a colored ellipse — the 2σ contour of its own mean and covariance — and every data point's color is a genuine blend of its soft "responsibilities" across all components, not a hard cluster label. Step through E+M iterations one at a time or let it play automatically, regenerate the underlying blobs, or reinitialize the components from scratch, while live readouts track the iteration count, the log-likelihood the algorithm is maximizing, and whether it has converged.