Id: arxiv/2611.25147 Url: https://arxiv.org/abs/2611.25147 Tags: quant-ph, math.NA Score: 0 Title: Accelerating Diffusion Maps on a Quantum Computer: Spectral Filtering and Galerkin PDE Solvers Domain: quant-ph Source: arxiv Authors: Ruizhe Qian, Xiantao Li Reviews: 0 Subject: quant-ph Category: research Comments: 0 Disputed: false Division: physics Mentions: 0 Discussed: 0 Published: 2026-11-18 Page: https://delphi.science/r/arxiv/2611.25147 Diffusion maps generalize principal component analysis (PCA) to nonlinear manifolds, enabling spectral dimensionality reduction, geometric data analysis, and elliptic PDE approximation on point clouds. For large datasets of size $N$, however, classical matrix operations incur prohibitive $\operatorname{poly}(N)$ scaling, even when extracting low-dimensional spectral features or targeted scalar observables. We introduce a quantum algorithmic framework that bypasses explicit graph-Laplacian assembly by exploiting the symmetric factorization $G = A^T A$ directly at the block-encoding level.…