Distributed Multi-Task Learning with Shared Representation
arXiv:1603.02185
Abstract
We study the problem of distributed multi-task learning with shared representation, where each machine aims to learn a separate, but related, task in an unknown shared low-dimensional subspaces, i.e. when the predictor matrix has low rank. We consider a setting where each task is handled by a different machine, with samples for the task available locally on the machine, and study communication-efficient methods for exploiting the shared structure.
References in corpus (7)
- Natural Language Processing (almost) from Scratch
- Communication Efficient Distributed Optimization using an Approximate Newton-type Method
- On the Linear Convergence of the Alternating Direction Method of Multipliers
- On the Global Linear Convergence of Frank-Wolfe Optimization Variants
- Communication-Efficient Distributed Dual Coordinate Ascent
- Communication-efficient sparse regression: a one-shot approach
- Distributed Multitask Learning
Cited by in corpus (7)
- Federated Multi-Task Learning
- Distributed Stochastic Multi-Task Learning with Graph Regularization
- Decentralized Multi-Task Learning Based on Extreme Learning Machines
- Fully Decentralized Joint Learning of Personalized Models and Collaboration Graphs
- Recovery of simultaneous low rank and two-way sparse coefficient matrices, a nonconvex approach
- Distributed Multi-Task Relationship Learning
- Distributed Primal-Dual Optimization for Online Multi-Task Learning