On the exponential value of labeled samples
Pattern Recognition Letters
Ongoing Learning for Supervised Pattern Recognition
SIBGRAPI '01 Proceedings of the 14th Brazilian Symposium on Computer Graphics and Image Processing
Learning from Labeled and Unlabeled Data using Graph Mincuts
ICML '01 Proceedings of the Eighteenth International Conference on Machine Learning
Partially supervised clustering for image segmentation
Pattern Recognition
A stochastic approach to wilson's editing algorithm
IbPRIA'05 Proceedings of the Second Iberian conference on Pattern Recognition and Image Analysis - Volume Part II
Non parametric local density-based clustering for multimodal overlapping distributions
IDEAL'06 Proceedings of the 7th international conference on Intelligent Data Engineering and Automated Learning
| Hi-index | 0.01 |
The performance of supervised learners depends on the presence of a relatively large labeled sample. This paper proposes an automatic ongoing learning system, which is able to incorporate new knowledge from the experience obtained when classifying new objects and correspondingly, to improve the efficiency of the system. We employ a stochastic rule for classifying and editing, along with a condensing algorithm based on local density to forget superfluous data (and control the sample size). The effectiveness of the algorithm is experimentally evaluated using a number of data sets taken from the UCI Machine Learning Database Repository.