Manifold Learning via Manifold Deflation
This work addresses robustness issues in manifold learning for data visualization and interpretation, but it appears incremental as it builds on existing spectral methods with a new estimator.
The paper tackles the problem of nonlinear dimensionality reduction methods failing on simple manifolds due to issues like noise and boundary bias, and it introduces a manifold deflation method that iteratively eliminates dimensions from a differential operator, showing empirical recovery of novel embeddings on real-world and synthetic datasets.
Nonlinear dimensionality reduction methods provide a valuable means to visualize and interpret high-dimensional data. However, many popular methods can fail dramatically, even on simple two-dimensional manifolds, due to problems such as vulnerability to noise, repeated eigendirections, holes in convex bodies, and boundary bias. We derive an embedding method for Riemannian manifolds that iteratively uses single-coordinate estimates to eliminate dimensions from an underlying differential operator, thus "deflating" it. These differential operators have been shown to characterize any local, spectral dimensionality reduction method. The key to our method is a novel, incremental tangent space estimator that incorporates global structure as coordinates are added. We prove its consistency when the coordinates converge to true coordinates. Empirically, we show our algorithm recovers novel and interesting embeddings on real-world and synthetic datasets.