Spectral Methods for Regularization in Learning Theory

By Ernesto De Vito, Lorenzo Rosasco, Alessandro Verri. 2006

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

Kernel Mean Embedding of Distributions: A Review and Beyonds (2016, methodology): “See, e.g., Engl et al. (1996), Bauer et al. (2007), De Vito et al. (2006), Muandet et al. (2014b) for detail of each algorithm.”
Kernel Mean Estimation via Spectral Filtering (2014, methodology): “The step-size η does not play any role for shrinkage [16], so we use the fixed step-size η = 1/κ(2) throughout.”
Density Estimation in Infinite Dimensional Exponential Families (2013, methodology): “In the statistical context, [29, 3] have used the ideas from [12] in non-parametric regression for learning a square integrable function from finite samples through regularization in RKHS.”
Optimal Rates for Regularization Operators in Learning Theory (2006, methodology): “In the context of learning theory a review of these techniques can be found in [10] and [1].”
Kernel ridge vs. principal component regression: Minimax bounds and the qualification of regularization operators (2017, methodology): “In particular, KPCR (with infinite qualification) can adapt to source conditions with arbitrary ζ ≥ 0; on the other hand, KRR satisfies the conditions of Corollary 1 only when ζ ≤ 1, because KRR has qualification 1.”
On regularization algorithms in learning theory (2007, methodology): “The first attempt to resolve this discrepancy has been made in De Vito et al. (2005b); Yao et al. (2005); Rosasco et al. (2005), where the estimates of the Lipschitz constants of functions generating regularization methods have been used for obtaining error bounds.”
Spectral Algorithms for Supervised Learning (2008, methodology): “See De Vito, Rosasco, and Verri (2005), Bauer et al. (2007), and Caponnetto (2006), for arbitrary ﬁlters.”