Gérard Villermain Lecolier

Learn More
We use learning methods to build classiÿers in using a set of training samples. A classiÿer is capable to assign a new sample into one of the diierent learnt classes. In a non-stationary work environment, a classiÿer must be retrained every time a new sample is classiÿed, to obtain new knowledge from it. This is an impractical solution since it requires the(More)
  • 1