Energy-constrained discriminant analysis
Dimensionality reduction algorithms have become an indispensable tool for working with high-dimensional data in classification. Linear discriminant analysis (LDA) is a popular analysis technique used to project high-dimensional data into a lower-dimensional space while maximizing class separability....
| Main Authors: | , , |
|---|---|
| Other Authors: | |
| Format: | Article |
| Language: | en_US |
| Published: |
Institute of Electrical and Electronics Engineers
2010
|
| Online Access: | http://hdl.handle.net/1721.1/60231 |
| _version_ | 1826203285290221568 |
|---|---|
| author | Philips, Scott M. Berisha, Visar Spanias, Andreas |
| author2 | Lincoln Laboratory |
| author_facet | Lincoln Laboratory Philips, Scott M. Berisha, Visar Spanias, Andreas |
| author_sort | Philips, Scott M. |
| collection | MIT |
| description | Dimensionality reduction algorithms have become an indispensable tool for working with high-dimensional data in classification. Linear discriminant analysis (LDA) is a popular analysis technique used to project high-dimensional data into a lower-dimensional space while maximizing class separability. Although this technique is widely used in many applications, it suffers from overfitting when the number of training examples is on the same order as the dimension of the original data space. When overfitting occurs, the direction of the LDA solution can be dominated by low-energy noise and therefore the solution becomes non-robust to unseen data. In this paper, we propose a novel algorithm, energy-constrained discriminant analysis (ECDA), that overcomes the limitations of LDA by finding lower dimensional projections that maximize inter-class separability, while also preserving signal energy. Our results show that the proposed technique results in higher classification rates when compared to comparable methods. The results are given in terms of SAR image classification, however the algorithm is broadly applicable and can be generalized to any classification problem. |
| first_indexed | 2024-09-23T12:34:14Z |
| format | Article |
| id | mit-1721.1/60231 |
| institution | Massachusetts Institute of Technology |
| language | en_US |
| last_indexed | 2024-09-23T12:34:14Z |
| publishDate | 2010 |
| publisher | Institute of Electrical and Electronics Engineers |
| record_format | dspace |
| spelling | mit-1721.1/602312022-10-01T09:50:12Z Energy-constrained discriminant analysis Philips, Scott M. Berisha, Visar Spanias, Andreas Lincoln Laboratory Philips, Scott M. Philips, Scott M. Berisha, Visar Dimensionality reduction algorithms have become an indispensable tool for working with high-dimensional data in classification. Linear discriminant analysis (LDA) is a popular analysis technique used to project high-dimensional data into a lower-dimensional space while maximizing class separability. Although this technique is widely used in many applications, it suffers from overfitting when the number of training examples is on the same order as the dimension of the original data space. When overfitting occurs, the direction of the LDA solution can be dominated by low-energy noise and therefore the solution becomes non-robust to unseen data. In this paper, we propose a novel algorithm, energy-constrained discriminant analysis (ECDA), that overcomes the limitations of LDA by finding lower dimensional projections that maximize inter-class separability, while also preserving signal energy. Our results show that the proposed technique results in higher classification rates when compared to comparable methods. The results are given in terms of SAR image classification, however the algorithm is broadly applicable and can be generalized to any classification problem. United States. Defense Advanced Research Projects Agency (Air Force Contract FA8721-05-C-0002) 2010-12-08T18:23:46Z 2010-12-08T18:23:46Z 2009-05 2009-04 Article http://purl.org/eprint/type/ConferencePaper 978-1-4244-2353-8 1520-6149 INSPEC Accession Number: 10701068 http://hdl.handle.net/1721.1/60231 Philips, S., V. Berisha, and A. Spanias. “Energy-constrained discriminant analysis.” Acoustics, Speech and Signal Processing, 2009. ICASSP 2009. IEEE International Conference on. 2009. 3281-3284. © 2009 IEEE. en_US http://dx.doi.org/10.1109/ICASSP.2009.4960325 IEEE International Conference on Acoustics, Speech and Signal Processing, 2009. ICASSP 2009 Article is made available in accordance with the publisher's policy and may be subject to US copyright law. Please refer to the publisher's site for terms of use. application/pdf Institute of Electrical and Electronics Engineers IEEE |
| spellingShingle | Philips, Scott M. Berisha, Visar Spanias, Andreas Energy-constrained discriminant analysis |
| title | Energy-constrained discriminant analysis |
| title_full | Energy-constrained discriminant analysis |
| title_fullStr | Energy-constrained discriminant analysis |
| title_full_unstemmed | Energy-constrained discriminant analysis |
| title_short | Energy-constrained discriminant analysis |
| title_sort | energy constrained discriminant analysis |
| url | http://hdl.handle.net/1721.1/60231 |
| work_keys_str_mv | AT philipsscottm energyconstraineddiscriminantanalysis AT berishavisar energyconstraineddiscriminantanalysis AT spaniasandreas energyconstraineddiscriminantanalysis |