Laplacian Eigenmaps from Sparse, Noisy Similarity Measurements
arXiv:1603.03972 · doi:10.1109/TSP.2016.2645517
Abstract
Manifold learning and dimensionality reduction techniques are ubiquitous in science and engineering, but can be computationally expensive procedures when applied to large data sets or when similarities are expensive to compute. To date, little work has been done to investigate the tradeoff between computational resources and the quality of learned representations. We present both theoretical and experimental explorations of this question. In particular, we consider Laplacian eigenmaps embeddings based on a kernel matrix, and explore how the embeddings behave when this kernel matrix is corrupted by occlusion and noise. Our main theoretical result shows that under modest noise and occlusion assumptions, we can (with high probability) recover a good approximation to the Laplacian eigenmaps embedding based on the uncorrupted kernel matrix. Our results also show how regularization can aid this approximation. Experimentally, we explore the effects of noise and occlusion on Laplacian eigenmaps embeddings of two real-world data sets, one from speech processing and one from neuroscience, as well as a synthetic data set.
References in corpus (11)
- Kernel methods in machine learning
- User-friendly tail bounds for sums of random matrices
- Spectral clustering and the high-dimensional stochastic blockmodel
- Matrix estimation by Universal Singular Value Thresholding
- Pseudo-likelihood methods for community detection in large sparse networks
- Regularized Spectral Clustering under the Degree-Corrected Stochastic Blockmodel
- Concentration of the adjacency matrix and of the Laplacian in random graphs with independent edges
- Localization from Incomplete Noisy Distance Measurements
- Finite metric spaces--combinatorics, geometry and algorithms
- Random perturbation of low rank matrices: Improving classical bounds
- A Joint Graph Inference Case Study: the C.elegans Chemical and Electrical Connectomes