On Spectral Learning

By Andreas Argyriou, Charles A. Micchelli, Massimiliano Pontil. 2010

310 citations recorded by the source corpus.

Subfields: Computational Mechanics, Computational Theory and Mathematics, Mathematical Physics

How researchers see this paper

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

In Computational Mechanics

In Computational Theory and Mathematics

In Mathematical Physics

Citations in the corpus

A Survey on Multi-Task Learning (2017, background): “[228], [229] discuss conditions for regularized MTL algorithms that representer theorems hold.”
A Comprehensive Survey on Transfer Learning (2019, background): “Instead of seeking the discrete solution of the indicator vector, the solution is relaxed to be continuous, and the eigen-system problem corresponding to the objective function is solved to construct the aligned spectral features [111].”
Unsupervised Deep Learning by Neighbourhood Discovery (2019, background): “Broadly, AND relates to constrained clustering (Wagstaff et al., 2001; Kamvar et al., 2003; Zhu et al., 2013; 2016) if considering our neighbourhood constraint as a form of pairwise supervision including must-link and cannot-link.”
A whole brain fMRI atlas generated via spatially constrained spectral clustering (2011, methodology): “…was chosen for clustering due to its robustness to outliers [Shi and Malik, 2000; von Luxburg, 2007], favorable performance when compared to other algorithms [Shen et al., 2010], and the simplicity with which it can be adapted to incorporate constraints [Kamvar et al., 2003; Xu et al., 2005].”
Semi-Supervised Kernel Mean Shift Clustering (2014, methodology): “In [23], this method was extended to incorporate weak supervision by updating the computed affinity matrix for the specified constraint points.”
An Active Learning Approach to Coreference Resolution (2015, background): “However, we should point out that our work also has connections with constrained Spectral Clustering [Kamvar et al., 2003], Constrained K-means [Wagstaff et al., 2001], Semi-supervised clustering [Kulis et al., 2005].”
Learning Must-Link Constraints for Video Segmentation Based on Spectral Clustering (2014, background): “[24], which first considered constrained spectral clustering); ii.”
Nuclear norm penalization and optimal rates for noisy low rank matrix completion (2010, methodology): “The most popular methods are based on penalized empirical risk minimization with nuclear norm penalty [4, 2, 3, 5, 7, 8, 9, 13, 18, 23].”
Semi-supervised graph clustering: a kernel approach (2005, methodology): “One spectral approach to semi-supervised clustering is the S PECTRAL - LEARNING algorithm of Kamvar et al. (2003): • Form the afﬁnity matrix A : the entries of A are assumed to be normalized between 0 and 1, where entry A ij is the similarity between points i and j .”
Constrained 1-Spectral Clustering (2012, methodology): “We compare our method against the following four related constrained clustering approaches: Spectral Learning (SL) (Kamvar et al., 2003), Flexible Constrained Spectral Clustering (CSP) (Wang & David-son, 2010), Constrained Clustering via Spectral Regu-larization (CCSR) (Li et al., 2009) and…”
What you saw is not what you get: Domain adaptation using asymmetric kernel transforms (2011, background): “We also note connections to the work of [1], which derives representer theorems for various matrix learning problems.”
Active co-analysis of a set of shapes (2012, methodology): “Many existing techniques directly modify the similarity matrix according to the given constraints and then perform spectral clustering [Kamvar et al. 2003; Kulis et al. 2005; Lu and Carreira-Perpinán 2008].”
Estimation of high-dimensional low-rank matrices (2009, background): “Argyriou et al. (2008), Ar-gyriou, Evgeniou and Pontil (2008), Argyriou, Micchelli and Pontil (2010), Bach (2008), Abernethy et al. (2009) discussed connections of (2.5) to other related minimization problems, along with characterizations of the solutions and computational issues, mainly focusing…”
Harnessing Lab Knowledge for Real-World Action Recognition (2014, methodology): “Further inspired by Argyriou et al. (2010), Nie et al. (2012), we propose to use the general Schatten p-norm to approximate the matrix rank.”
Spectral Clustering for Speech Separation (2009, background): “Several recent papers have considered ways to alleviate this burden by incorporating prior knowledge into the metric, either in the setting of K-means clustering (Wagstaff et al., 2001, Xing et al., 2003, Bar-Hillel et al., 2003) or spectral clustering (Yu and Shi, 2002, Kamvar et al., 2003).”