Year: 2009
Type: conference-paper
Abstract: A method for training overlapping feed-forward networks on analogous tasks is extended and analyzed. The learning dynamics of simultaneous (interlaced) training of similar tasks interact at the shared connections of the networks. The output of one network in re-sponse to a stimulus to the other network can be interpreted as an analogical inference. In a similar fashion, the networks can be explicitly trained to map specific items in one domain to specifc items in the other domain. The method has been applied to spatial tasks in a simple environments and to tree structures.
Authors: Paul Munro, Yefei Peng
Institution: University of Pittsburgh
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W2264866997