Improved guarantees and a multiple-descent curve for Column Subset Selection and the Nyström method
This work addresses fundamental limitations in constructing efficient low-rank approximations for large datasets in machine learning and scientific computing, offering theoretical insights and practical implications for data subset selection.
The paper tackles the Column Subset Selection Problem (CSSP) and Nyström method for low-rank approximations by developing techniques that exploit spectral properties to achieve improved approximation guarantees beyond worst-case analysis, particularly for datasets with polynomial or exponential singular value decay, and reveals a multiple-descent curve phenomenon in the approximation factor as a function of subset size k.
The Column Subset Selection Problem (CSSP) and the Nyström method are among the leading tools for constructing small low-rank approximations of large datasets in machine learning and scientific computing. A fundamental question in this area is: how well can a data subset of size k compete with the best rank k approximation? We develop techniques which exploit spectral properties of the data matrix to obtain improved approximation guarantees which go beyond the standard worst-case analysis. Our approach leads to significantly better bounds for datasets with known rates of singular value decay, e.g., polynomial or exponential decay. Our analysis also reveals an intriguing phenomenon: the approximation factor as a function of k may exhibit multiple peaks and valleys, which we call a multiple-descent curve. A lower bound we establish shows that this behavior is not an artifact of our analysis, but rather it is an inherent property of the CSSP and Nyström tasks. Finally, using the example of a radial basis function (RBF) kernel, we show that both our improved bounds and the multiple-descent curve can be observed on real datasets simply by varying the RBF parameter.