PROSPECTS FOR AUTOMATIC RECODING OF INPUTS IN CONNECTIONIST LEARNING

Authors
SZILAS N SHULTZ TR
Citation
N. Szilas et Tr. Shultz, PROSPECTS FOR AUTOMATIC RECODING OF INPUTS IN CONNECTIONIST LEARNING, Behavioral and brain sciences, 20(1), 1997, pp. 81
Citations number
9
Categorie Soggetti
Psychology,"Psychology, Biological",Neurosciences,"Behavioral Sciences
Journal title
Behavioral and brain sciences → ACNP
ISSN journal
0140525X
Volume
20
Issue
1
Year of publication
1997
Database
ISI
SICI code
0140-525X(1997)20:1<81:PFAROI>2.0.ZU;2-R
Abstract
Clark & Thornton present the well-established principle that recoding inputs can make learning easier. A useful goal would be to make such r ecoding automatic. We discuss some ways in which incrementality and tr ansfer in connectionist networks could attain this goal.