A Regularization Approach to Learning Task Relationships in Multitask Learning

By Yu Zhang, Dit-Yan Yeung. 2014

143 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Civil and Structural Engineering, Representation & Self-Supervised Learning

How researchers see this paper

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

In Artificial Intelligence

In Civil and Structural Engineering

In Representation & Self-Supervised Learning

Citations in the corpus

An Overview of Multi-task Learning (2018, methodology): “A regularized model called multi-task-relationship learning (MTRL) method is proposed in [44,45] by placing a matrix-variate normal prior on W : W ∼ MN ( 0 , I , (cid:6) ), where MN ( M , A , B ) denotes a matrix-variate normal distribution with M , A , B as the mean, row covariance and column…”
Wireless Network Intelligence at the Edge (2018, background): “For a given set of tasks, multitask learning (MTL) [170], [174] trains an NN so as to ensure high accuracy for multiple tasks by inserting a task correlation regularizer into the original loss function [174].”
A Survey on Multi-Task Learning (2017, methodology): “Different from [79], [81] which are Bayesian models, Zhang and Yeung [4], [5] propose a regularized multi-task model called multi-task relationship learning (MTRL) by placing a matrixvariate normal prior on W as”
Adaptive Auxiliary Task Weighting for Reinforcement Learning (2019, background): “task relationships to be known as a prior [25], or learned from the relationships among task specific parameters [26, 27].”
Reconciling meta-learning and continual learning with online mixtures of tasks (2018, background): “Moreover, explicitly estimating relatedness between tasks has the potential to aid the efﬁcacy of a meta-learning algorithm by modulating both positive and negative transfer [52, 59, 45, 61], and by identifying outlier tasks that require a more signiﬁcant degree of adaptation [56, 23].”
Multi-Task Multiple Kernel Relationship Learning (2016, methodology): “By taking the partial derivative of the objective function with respect to Ω and setting it zero, we get an analytical solution for Ω [29]:”
Smart Multitask Bregman Clustering and Multitask Kernel Clustering (2015, methodology): “The other way is to learn the intertask covariance matrix of the multivariate Gaussian prior which can model both positive and negative task correlations [Fei and Huan 2013; Saha et al. 2011; Zhang and Yeung 2010, 2012a, 2012b, 2014].”
Multi-Task Reinforcement Learning with Soft Modularization (2020, background): “Researchers also propose to explicitly model the similarity between gradients from different tasks [44, 4, 15, 19, 35, 6, 42, 12].”
Learning Multiple Tasks with Multilinear Relationship Networks (2015, methodology): “ds. We compare with a variety of methods: Single-Task Softmax Regression (STSR), Multi-Task Feature Learning (MTFL) [3], Robust Multi-Task Learning (RMTL) [5], Multi-Task Relationship Learning (MTRL) [11, 13], and Multi-Task Deep Convolutional Neural Network (MTCNN) [18]. Speciﬁcally, LR is performe”
Multi-Task Feature Learning for Knowledge Graph Enhanced Recommendation (2019, background): “e of learning tasks further, including semi-supervised learning, active learning, unsupervised learning,and reinforcement learning. Our work can be seen as an asymmetric multi-task learning framework [37, 43, 44], in which we aim to utilize the connection between RS and KG to help improve their perf”
Adaptive Smoothed Online Multi-Task Learning (2016, background): “Most existing work in multi-task learning focuses on how to take advantage of these task relationships, either to share data directly [1] or to learn model parameters via cross-task regularization techniques [2, 3, 4].”