Composition of Conditional Random Fields for Transfer Learning

Composition of Conditional Random Fields for Transfer Learning (2005)

by Charles Sutton , Andrew McCallum

Venue:	PROCEEDINGS OF HLT/EMNLP
Citations:	21 - 1 self

User correction supplied by mph

Datum	Value	Source
TITLE	Composition of Conditional Random Fields for Transfer Learning	user correction
AUTHOR NAME	Charles Sutton	user correction
AUTHOR AFFIL	Department of Computer Science; University of Massachusetts	user correction
AUTHOR ADDR	Amherst, MA 01003	user correction
AUTHOR NAME	Andrew McCallum	user correction
AUTHOR AFFIL	Department of Computer Science; University of Massachusetts	user correction
AUTHOR ADDR	Amherst, MA 01003	user correction
ABSTRACT	Many learning tasks have subtasks for which much training data exists. Therefore, we want to transfer learning from the old, generalpurpose subtask to a more specific new task, for which there is often less data. While work in transfer learning often considers how the old task should affect learning on the new task, in this paper we show that it helps to take into account how the new task affects the old. Specifically, we perform joint decoding of separately-trained sequence models, preserving uncertainty between the tasks and allowing information from the new task to affect predictions on the old task. On two standard text data sets, we show that joint decoding outperforms cascaded decoding.	user correction
YEAR	2005	user correction
VENUE	PROCEEDINGS OF HLT/EMNLP	user correction
CITATIONS	21 found	ParsCit 1.0