ML Anthology

Cross-Domain Knowledge Transfer Using Structured Representations

Samarth Swarup, Sylvian R. Ray
AAAI 2006 pp. 506-511
/aaai/2006/swarup2006aaai-cross/

Abstract

Previous work in knowledge transfer in machine learn-ing has been restricted to tasks in a single domain. How-ever, evidence from psychology and neuroscience sug-gests that humans are capable of transferring knowl-edge across domains. We present here a novel learn-ing method, based on neuroevolution, for transferring knowledge across domains. We use many-layered, sparsely-connected neural networks in order to learn a structural representation of tasks. Then we mine fre-quent sub-graphs in order to discover sub-networks that are useful for multiple tasks. These sub-networks are then used as primitives for speeding up the learning of subsequent related tasks, which may be in different do-mains.

PDF AAAI Semantic Scholar

Cite

Text

Swarup and Ray. "Cross-Domain Knowledge Transfer Using Structured Representations." AAAI Conference on Artificial Intelligence, 2006.

Markdown

[Swarup and Ray. "Cross-Domain Knowledge Transfer Using Structured Representations." AAAI Conference on Artificial Intelligence, 2006.](https://mlanthology.org/aaai/2006/swarup2006aaai-cross/)

BibTeX

@inproceedings{swarup2006aaai-cross,
  title     = {{Cross-Domain Knowledge Transfer Using Structured Representations}},
  author    = {Swarup, Samarth and Ray, Sylvian R.},
  booktitle = {AAAI Conference on Artificial Intelligence},
  year      = {2006},
  pages     = {506-511},
  url       = {https://mlanthology.org/aaai/2006/swarup2006aaai-cross/}
}