SIGNAL PROCESSING LETTERS 1 On-Line Learning for Active Pattern

  title={SIGNAL PROCESSING LETTERS 1 On-Line Learning for Active Pattern},
  author={R Park},
| An adaptive on-line learning method is presented to faciliate pattern classiication using active sampling to identify optimal decision boundary for a stochas-tic oracle with minimum number of training samples. The strategy of sampling at the current estimate of the decision boundary is shown to be optimal compared to random sampling in the sense that the probability of convergence toward the true decision boundary at each step is maximized, ooering theoretical justiication on the popular… CONTINUE READING


Publications referenced by this paper.
Showing 1-9 of 9 references

Cohn , \ Neural network exploration using optimal experiment design "

A. David
Advances in Neural Information ProcessingSystems • 1994

Incremental learning with and without queries in binary choice problems,

Yoshiyuki Kabashima, Shigeru Shinomoto
Proc. International Joint Conference on Neural Networks, • 1993
View 1 Excerpt

, \ Active selection of trainingexamples for network learning in noiseless environments , " Tech

Mark Plutowski, Halbert White

Active selection of training examples for network learning in noiseless environments,

Mark Plutowski, Halbert White
Tech. Rep. CS91-180, • 1990
View 2 Excerpts

Angluin , \ Queries and concept learning


Queries and concept learning

Machine Learning • 1987
View 1 Excerpt

Similar Papers

Loading similar papers…