Learning One More Thing

Sebastian Thrun and Tom Mitchell

Most research on machine learning has focused on scenarios in which a learner faces a single, isolated learning task. The lifelong learning framework assumes that the learner encounters a multitude of related learning tasks over its lifetime, providing the opportunity for the transfer of knowledge among these. This paper studies lifelong learning in the context of binary classification. It presents the invariance approach, in which knowledge is transferred via a learned model of the invariances of the domain. Results on learning to recognize objects from color images demonstrate superior generalization capabilities if invariances are learned and used to bias subsequent learning.

Click here to obtain the full paper (305633 bytes).

@INPROCEEDINGS{Thrun95e,
  AUTHOR         = {S. Thrun and T. Mitchell},
  YEAR           = {1995},
  TITLE          = {Learning One More Thing},
  BOOKTITLE      = {Proceedings of the Fourteenth International Joint 
                    Conference on Artificial Intelligence (IJCAI)},
  PUBLISHER      = {Morgan Kaufmann},
  ADDRESS        = {San Mateo, CA}
}