Spectral Clustering via Fiedler Vector

Graph partitioning using the algebraic connectivity eigenvector

Graph colored by cluster

Fiedler vector components

λ₂ = — | Cut ratio = — | Accuracy = —
The normalized graph Laplacian L = D⁻¹/²(D−A)D⁻¹/² has eigenvalues 0 = λ₁ ≤ λ₂ ≤ … ≤ λ_n. The Fiedler value λ₂ = algebraic connectivity measures how well-connected the graph is. The Fiedler vector v₂ (corresponding eigenvector) encodes the optimal graph bisection: partition by sign of v₂(i). For k clusters, use k smallest non-zero eigenvectors (k-means on eigenvectors). Computed via power iteration on the deflated Laplacian.