The Deep Belief and Self-Organizing Neural Network as a Semi-Supervised Classification Method for Hyperspectral Data
Hyperspectral data is not linearly separable, and it has a high characteristic dimension. This paper proposes a new algorithm that combines a deep belief network based on the Boltzmann machine with a self-organizing neural network. The primary features of the hyperspectral image are extracted with a...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2017-11-01
|
| Series: | Applied Sciences |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2076-3417/7/12/1212 |
| _version_ | 1818496589006635008 |
|---|---|
| author | Wei Lan Qingjian Li Nan Yu Quanxin Wang Suling Jia Ke Li |
| author_facet | Wei Lan Qingjian Li Nan Yu Quanxin Wang Suling Jia Ke Li |
| author_sort | Wei Lan |
| collection | DOAJ |
| description | Hyperspectral data is not linearly separable, and it has a high characteristic dimension. This paper proposes a new algorithm that combines a deep belief network based on the Boltzmann machine with a self-organizing neural network. The primary features of the hyperspectral image are extracted with a deep belief network. The weights of the network are fine-tuned using the labeled sample. Feature vectors extracted by the deep belief network are classified by a self-organizing neural network. The method reduces the spectral dimension of the data while preserving the large amount of original information in the data. The method overcomes the long training time required when using self-organizing neural networks for clustering, as well as the training difficulties of Deep Belief Networks (DBN) when the labeled sample size is small, thereby improving the accuracy and robustness of the semi-supervised classification. Simulation results show that the structure of the network can achieve higher classification accuracy when the labeled sample is deficient. |
| first_indexed | 2024-12-10T18:34:55Z |
| format | Article |
| id | doaj.art-c16f8b71912f402a8193a5db05aa1c47 |
| institution | Directory Open Access Journal |
| issn | 2076-3417 |
| language | English |
| last_indexed | 2024-12-10T18:34:55Z |
| publishDate | 2017-11-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Applied Sciences |
| spelling | doaj.art-c16f8b71912f402a8193a5db05aa1c472022-12-22T01:37:51ZengMDPI AGApplied Sciences2076-34172017-11-01712121210.3390/app7121212app7121212The Deep Belief and Self-Organizing Neural Network as a Semi-Supervised Classification Method for Hyperspectral DataWei Lan0Qingjian Li1Nan Yu2Quanxin Wang3Suling Jia4Ke Li5School of Economics and Management, Beihang University, Beijing 100191, ChinaFundamental Science on Ergonomics and Environment Control Laboratory, School of Aeronautics Science and Engineering, Beihang University, Beijing 100191, ChinaFundamental Science on Ergonomics and Environment Control Laboratory, School of Aeronautics Science and Engineering, Beihang University, Beijing 100191, ChinaFundamental Science on Ergonomics and Environment Control Laboratory, School of Aeronautics Science and Engineering, Beihang University, Beijing 100191, ChinaSchool of Economics and Management, Beihang University, Beijing 100191, ChinaFundamental Science on Ergonomics and Environment Control Laboratory, School of Aeronautics Science and Engineering, Beihang University, Beijing 100191, ChinaHyperspectral data is not linearly separable, and it has a high characteristic dimension. This paper proposes a new algorithm that combines a deep belief network based on the Boltzmann machine with a self-organizing neural network. The primary features of the hyperspectral image are extracted with a deep belief network. The weights of the network are fine-tuned using the labeled sample. Feature vectors extracted by the deep belief network are classified by a self-organizing neural network. The method reduces the spectral dimension of the data while preserving the large amount of original information in the data. The method overcomes the long training time required when using self-organizing neural networks for clustering, as well as the training difficulties of Deep Belief Networks (DBN) when the labeled sample size is small, thereby improving the accuracy and robustness of the semi-supervised classification. Simulation results show that the structure of the network can achieve higher classification accuracy when the labeled sample is deficient.https://www.mdpi.com/2076-3417/7/12/1212DBNdata compressionhyperspectralSOMpattern classification |
| spellingShingle | Wei Lan Qingjian Li Nan Yu Quanxin Wang Suling Jia Ke Li The Deep Belief and Self-Organizing Neural Network as a Semi-Supervised Classification Method for Hyperspectral Data Applied Sciences DBN data compression hyperspectral SOM pattern classification |
| title | The Deep Belief and Self-Organizing Neural Network as a Semi-Supervised Classification Method for Hyperspectral Data |
| title_full | The Deep Belief and Self-Organizing Neural Network as a Semi-Supervised Classification Method for Hyperspectral Data |
| title_fullStr | The Deep Belief and Self-Organizing Neural Network as a Semi-Supervised Classification Method for Hyperspectral Data |
| title_full_unstemmed | The Deep Belief and Self-Organizing Neural Network as a Semi-Supervised Classification Method for Hyperspectral Data |
| title_short | The Deep Belief and Self-Organizing Neural Network as a Semi-Supervised Classification Method for Hyperspectral Data |
| title_sort | deep belief and self organizing neural network as a semi supervised classification method for hyperspectral data |
| topic | DBN data compression hyperspectral SOM pattern classification |
| url | https://www.mdpi.com/2076-3417/7/12/1212 |
| work_keys_str_mv | AT weilan thedeepbeliefandselforganizingneuralnetworkasasemisupervisedclassificationmethodforhyperspectraldata AT qingjianli thedeepbeliefandselforganizingneuralnetworkasasemisupervisedclassificationmethodforhyperspectraldata AT nanyu thedeepbeliefandselforganizingneuralnetworkasasemisupervisedclassificationmethodforhyperspectraldata AT quanxinwang thedeepbeliefandselforganizingneuralnetworkasasemisupervisedclassificationmethodforhyperspectraldata AT sulingjia thedeepbeliefandselforganizingneuralnetworkasasemisupervisedclassificationmethodforhyperspectraldata AT keli thedeepbeliefandselforganizingneuralnetworkasasemisupervisedclassificationmethodforhyperspectraldata AT weilan deepbeliefandselforganizingneuralnetworkasasemisupervisedclassificationmethodforhyperspectraldata AT qingjianli deepbeliefandselforganizingneuralnetworkasasemisupervisedclassificationmethodforhyperspectraldata AT nanyu deepbeliefandselforganizingneuralnetworkasasemisupervisedclassificationmethodforhyperspectraldata AT quanxinwang deepbeliefandselforganizingneuralnetworkasasemisupervisedclassificationmethodforhyperspectraldata AT sulingjia deepbeliefandselforganizingneuralnetworkasasemisupervisedclassificationmethodforhyperspectraldata AT keli deepbeliefandselforganizingneuralnetworkasasemisupervisedclassificationmethodforhyperspectraldata |