Skip to Main content Skip to Navigation
Reports

Learning the Graph of Relations Among Multiple Tasks

Abstract : We propose multitask Laplacian learning, a new method for jointly learning clusters of closely related tasks. Unlike standard multitask methodologies, the graph of relations among the tasks is not assumed to be known a priori, but is learned by the multitask Laplacian algorithm. The algorithm builds on kernel based methods and exploits an optimization approach for learning a continuously parameterized kernel. It involves solving a semidefinite program of a particu- lar type, for which we develop an algorithm based on Douglas-Rachford split- ting methods. Multitask Laplacian learning can find application in many cases in which tasks are related with each other to varying degrees, some strongly, oth- ers weakly. Our experiments highlight such cases in which multitask Laplacian learning outperforms independent learning of tasks and state of the art multitask learning methods. In addition, they demonstrate that our algorithm partitions the tasks into clusters each of which contains well correlated tasks.
Document type :
Reports
Complete list of metadata

Cited literature [27 references]  Display  Hide  Download

https://hal.inria.fr/hal-00940321
Contributor : Puneet Kumar Dokania <>
Submitted on : Monday, February 3, 2014 - 1:37:45 PM
Last modification on : Thursday, October 15, 2020 - 3:28:19 PM
Long-term archiving on: : Sunday, April 9, 2017 - 5:03:54 AM

File

MTGraph.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-00940321, version 1

Collections

Citation

Andreas Argyriou, Stéphan Clémençon, Ruocong Zhang. Learning the Graph of Relations Among Multiple Tasks. [Research Report] 2013. ⟨hal-00940321⟩

Share

Metrics

Record views

1256

Files downloads

874