Bounds for Linear Multi-Task Learning

By Andreas Maurer. 2006

178 citations recorded by the source corpus.

How researchers see this paper

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

In Statistical Learning Theory

Citations in the corpus

An Overview of Multi-task Learning (2018, neutral): “Thenthereare manystudiestoanalyzegeneralizationboundsofdif-ferentMTLapproaches,includinge.g. [7,134]for thefeaturetransformapproach,[135]forthefea-tureselectionapproach,[24,135–138]forthelow-rankapproach,[136]forthetask-relationlearning approach,and[138]forthedirtyapproach.”
Semi-Cyclic Stochastic Gradient Descent (2019, background): “Evgeniou & Pontil, 2004), or graph-based relatedness (e.g. Evgeniou et al., 2005; Maurer, 2006).”
A Survey on Multi-Task Learning (2017, methodology): “In [222], Maurer analyzes the generalization bound of a MTL method based on the Rademacher complexity, which learns a common feature transformation for all the tasks and is a special case of problem (11) without U, and then in [223] further analyzes generalization bounds of regularized MTL models ba”
Model Zoo: A Growing Brain That Learns Continually (2021, background): “Theoretical work on learning from multiple tasks Works such as Baxter (2000); Maurer (2006 2008) who show that learning diverse tasks requires a larger feature generator and, thereby, more samples; we discuss this in §2.2.”
Joint covariate selection and joint subspace selection for multiple classification problems (2010, background): “The general problem of jointly estimating models from multiple, related data sets is often referred to as “transfer learning” or “multi-task learning” in the machine learning literature (Maurer 2006; Ben-David and Schuller-Borbely 2008; Argyriou et al. 2008; Jebara 2004; Evgeniou and Pontil 2004; To”
The Benefit of Multitask Representation Learning (2015, background): “Other statistical learning bounds are restricted to linear multitask learning such as (Cavallanti et al., 2010; Lounici et al., 2011; Maurer, 2006a,a).”
Sluice networks: Learning what to share between loosely related tasks (2017, background): “This approach to multi-task learning is easy to implement, reduces overﬁtting and is guaranteed to work for (certain types of) closely related tasks [2, 20].”
Excess risk bounds for multitask learning with trace norm regularization (2012, background): “Similar results are obtained in [21].”
Regularization Techniques for Learning with Matrices (2009, background): “This bound follows easily from Talagrand’s contraction inequality and Thm. 8 in Maurer [2006].”
Multi-Task Learning and Algorithmic Stability (2015, methodology): “Maurer (2006) uses the Rademacher complexity to analyze linear multi-task methods.”
On Spectral Learning (2010, methodology): “This occurs, for example, in multi-task learning and problems closely related to it (Abernethy et al., 2009; Argyriou et al., 2007a,b; Cand̀es and Recht, 2008; Cavallanti et al., 2008; Izenman, 1975; Maurer, 2006a,a; Srebro et al., 2005; Yuan et al., 2007, etc.).”
Local Rademacher Complexity-based Learning Guarantees for Multi-Task Learning (2016, methodology): “MTL learning guarantees are centered around the notion of (global) Rademacher complexities; notions that were introduced to the machine learning community by [7, 9, 34, 31, 30], and employed in the context of MTL by [45, 46, 26, 48, 47, 49].”
A Spectral Regularization Framework for Multi-Task Structure Learning (2007, background): “Recently, there has been renewed interest in the problem of multi-task learning, see [2, 4, 5, 14, 16, 19] and references therein.”
Convex multi-task feature learning (2008, background): “Results in [17, 35 ] may be useful for this purpose.”
Why and when can deep-but not shallow-networks avoid the curse of dimensionality: A review (2016, background): “Multi-task learning has theoretical advantages that depends on compositionality: the sample complexity of the problem can be signiﬁcantly lower (see [45] ).”