Generalized Robust Adaptive-Bandwidth Multi-View Manifold Learning in High Dimensions with Noise
This provides a practical and theoretically grounded solution for multiview sensor fusion in scientific and engineering applications, addressing heterogeneous noise and high-dimensional data.
The paper tackles the problem of integrating multiple noisy data sources in high-dimensional settings by proposing GRAB-MDM, a kernel-based diffusion geometry framework with view-dependent bandwidth selection, which provably recovers shared intrinsic structures and shows significant improvements in robustness and embedding quality over baselines.
Multiview datasets are common in scientific and engineering applications, yet existing fusion methods offer limited theoretical guarantees, particularly in the presence of heterogeneous and high-dimensional noise. We propose Generalized Robust Adaptive-Bandwidth Multiview Diffusion Maps (GRAB-MDM), a new kernel-based diffusion geometry framework for integrating multiple noisy data sources. The key innovation of GRAB-MDM is a {view}-dependent bandwidth selection strategy that adapts to the geometry and noise level of each view, enabling a stable and principled construction of multiview diffusion operators. Under a common-manifold model, we establish asymptotic convergence results and show that the adaptive bandwidths lead to provably robust recovery of the shared intrinsic structure, even when noise levels and sensor dimensions differ across views. Numerical experiments demonstrate that GRAB-MDM significantly improves robustness and embedding quality compared with fixed-bandwidth and equal-bandwidth baselines, and usually outperform existing algorithms. The proposed framework offers a practical and theoretically grounded solution for multiview sensor fusion in high-dimensional noisy environments.