Classification of Camellia species from 3 sections using leaf anatomical data with back-propagation neural networks and support vector machines

Wu JIANG; Billur Barshan ÖZAKTAŞ; Nitin MANTRI; Zhengming TAO; Hongfei LU

Article Details

Classification of Camellia species from 3 sections using leaf anatomical data with back-propagation neural networks and support vector machines

Authors:

Wu JIANG | Billur Barshan ÖZAKTAŞ | Nitin MANTRI | Zhengming TAO | Hongfei LU

Publisher:

TUBITAK

Journal:

Turkish Journal of Botany 2013 Vol.37 Issue 6, pp.1093-1103

Document Type:

Article

Publication Date:

5/2013

Document Language:

Turkish

Document ID (Org No.):

2000722554

Frequency:

Bimonthly

eISSN:

1303-6106

pISSN:

1300-008X

Abstract:

 

Leaf characteristics provide many useful clues for taxonomy. We used a back-propagation artificial neural network (BP-ANN) and C-support vector machines (C-SVMs) to classify 47 species from 3 sections of genus Camellia (16 from sect. Chrysanthae, 16 from sect. Tuberculata, and 15 from sect. Paracamellia). The classification model was constructed based on 7 leaf anatomy attributes including, area of adaxial epidermal cell, thickness of adaxial epidermal cell, thickness of palisade parenchyma, thickness of total leaf, thickness of spongy parenchyma, thickness of abaxial epidermal cell, and area of abaxial epidermal cell. Model parameters of C-SVM, comprising regularization parameter (C) and kernel parameter (g), were optimized by cross-validation. The best classification accuracy of the 3 Camellia sections was achieved by the radial basis function SVM classifier (with parameters C = 32, g = 0.13), as well as the sigmoid SVM classifier (with parameters C = 32, g = 0.13), which was up to 84.00% in the training set and 90.91% in the prediction set, respectively. Compared with BP-ANN, SVM yields slightly higher prediction accuracy, which indicates that it is feasible to accurately classify the 3 sections of Camellia using SVMs based on leaf anatomy data.

 

Leaf characteristics provide many useful clues for taxonomy. We used a back-propagation artificial neural network (BP-ANN) and C-support vector machines (C-SVMs) to classify 47 species from 3 sections of genus Camellia (16 from sect. Chrysanthae, 16 from sect. Tuberculata, and 15 from sect. Paracamellia). The classification model was constructed based on 7 leaf anatomy attributes including, area of adaxial epidermal cell, thickness of adaxial epidermal cell, thickness of palisade parenchyma, thickness of total leaf, thickness of spongy parenchyma, thickness of abaxial epidermal cell, and area of abaxial epidermal cell. Model parameters of C-SVM, comprising regularization parameter (C) and kernel parameter (g), were optimized by cross-validation. The best classification accuracy of the 3 Camellia sections was achieved by the radial basis function SVM classifier (with parameters C = 32, g = 0.13), as well as the sigmoid SVM classifier (with parameters C = 32, g = 0.13), which was up to 84.00% in the training set and 90.91% in the prediction set, respectively. Compared with BP-ANN, SVM yields slightly higher prediction accuracy, which indicates that it is feasible to accurately classify the 3 sections of Camellia using SVMs based on leaf anatomy data.

Article Details

Classification of Camellia species from 3 sections using leaf anatomical data with back-propagation neural networks and support vector machines

Authors: Wu JIANG | Billur Barshan ÖZAKTAŞ | Nitin MANTRI | Zhengming TAO | Hongfei LU

Publisher: TUBITAK

Journal: Turkish Journal of Botany 2013 Vol.37 Issue 6, pp.1093-1103

Document Type: Article

Publication Date: 5/2013

Document Language: Turkish

Document ID (Org No.): 2000722554

Frequency: Bimonthly

eISSN: 1303-6106

pISSN: 1300-008X

Issues

Authors:

Wu JIANG | Billur Barshan ÖZAKTAŞ | Nitin MANTRI | Zhengming TAO | Hongfei LU

Publisher:

TUBITAK

Journal:

Turkish Journal of Botany 2013 Vol.37 Issue 6, pp.1093-1103

Document Type:

Article

Publication Date:

5/2013

Document Language:

Turkish

Document ID (Org No.):

2000722554

Frequency:

Bimonthly

eISSN:

1303-6106

pISSN:

1300-008X