We give lower bounds on the reconstruction error for PCA, k-means clustering, and va_absrious sparse coding methods. It is shown that the two objectives of good data approximation and sparsity of the solution are incompatible if the data distribution is evasive in the sense that most of its mass lies away from any low dimensional subspace. We give closure properties and examples of evasive distributions and quantify the extent to which distributions of bounded support and bounded density are evasive.
CITATION STYLE
Maurer, A., Pontil, M., & Baldassarre, L. (2015). Lower bounds for sparse coding. In Measures of Complexity: Festschrift for Alexey Chervonenkis (pp. 359–370). Springer International Publishing. https://doi.org/10.1007/978-3-319-21852-6_24
Mendeley helps you to discover research relevant for your work.