L. Jacob, F. Bach, J.-P. Vert

Clustered Multi-Task Learning: A Convex Formulation
Clustered multi-task learning Poster W33

Multi-task learning

...

...

· · ·

Related learning problems. Little data for some of them. Penalize variance across functions.

·
· ·

Groups of related learning problems. Penalize variance within each group. Optimal groups?

Metho d

·

W ,E

min

Loss(W ) + (W , E ),  penalizes variance within clusters E .
E
discrete

· ·

Set of clusterings



convex relaxation.

Recovers original clustering structure, improves prediction accuracy.