Learning from labeled and unlabeled data
Authors:
- 25 June 2003
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 3 (10987576) , 2803-2808
Abstract
Due to the considerable time and expense required in labeling data, a challenge is to propose learning algorithms that can learn from a small amount of labeled data and a much larger amount of unlabeled data. In this paper, we propose one such algorithm which uses an evolutionary strategy to iteratively adjust the class membership of the patterns in the unlabeled sample. The iterative adjustment is done so that the class conditional distribution obtained from such a labeling allows a maximum a posteriori classification with minimum classification error on the labeled patterns. We detail the algorithm and provide results obtained by the proposed algorithm on 5 different datasets.Keywords
This publication has 8 references indexed in Scilit:
- Combining labeled and unlabeled data with co-training1998
- Unsupervised word sense disambiguation rivaling supervised methods1995
- On the exponential value of labeled samplesPattern Recognition Letters, 1995
- Genetic Algorithms + Data Structures = Evolution Programs1994
- Small sample results for a linear discriminant function estimated from a mixture of normal populationsJournal of Statistical Computation and Simulation, 1979
- The efficiency of a linear discriminant function based on unclassified initial samplesBiometrika, 1978
- Normal Discrimination with Unclassified ObservationsJournal of the American Statistical Association, 1978
- On Estimation of a Probability Density Function and ModeThe Annals of Mathematical Statistics, 1962