An Efficient Active Learning Algorithm with Knowledge Transfer for Hyperspectral Data Analysis
- 1 January 2008
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 1 (21536996) , I-52-52
Abstract
We propose an active learning algorithm with knowledge transfer for classification of hyperspectral remote sensing data. The proposed method is based on a previously proposed algorithm, but yields faster learning curves by adjusting distributions of labeled data differently for the old and the new data. With the proposed method, the classifier can effectively transfer its knowledge learned from one region to a spatially or temporally separated region whose spectral signature is different. Empirical evaluation of the proposed algorithm is performed for two different hyperspectal datasets.Keywords
This publication has 8 references indexed in Scilit:
- An Active Learning Approach to Hyperspectral Data ClassificationIEEE Transactions on Geoscience and Remote Sensing, 2008
- Boosting for transfer learning2007
- Investigation of the random forest framework for classification of hyperspectral dataIEEE Transactions on Geoscience and Remote Sensing, 2005
- Best-bases feature extraction algorithms for classification of hyperspectral dataIEEE Transactions on Geoscience and Remote Sensing, 2001
- Active Learning with Statistical ModelsJournal of Artificial Intelligence Research, 1996
- A Sequential Algorithm for Training Text Classifiers1994
- Information-Based Objective Functions for Active Data SelectionNeural Computation, 1992
- Query by committee1992