DAGSVM vs. DAGKNN: An Experimental Case Study with Benthic Macroinvertebrate Dataset

In this paper we examined the suitability of the Directed Acyclic Graph Support Vector Machine (DAGSVM) and Directed Acyclic Graph k-Nearest Neighbour (DAGKNN) method in classification of the benthic macroinvertebrate samples. We divided our 50 species dataset into five ten species groups according to their group sizes. We performed extensive experimental tests with every group, where DAGSVM was tested with seven kernel functions and DAGKNN with four measures. Feature selection was made by the scatter method [8]. Results showed that the quadratic and RBF kernel functions were the best ones and in the case of DAGKNN all measures produced quite similar results. Generally, the DAGSVM gained higher accuracies than DAGKNN, but still DAGKNN is a respectable option in benthic macroinvertebrate classification.

[1]  Corinna Cortes,et al.  Support-Vector Networks , 1995, Machine Learning.

[2]  Martti Juhola,et al.  Automated Benthic Macroinvertebrate Identification with Decision Acyclic Graph Support Vector Machines , 2011 .

[3]  Thomas G. Dietterich,et al.  Automated processing and identification of benthic invertebrate samples , 2010, Journal of the North American Benthological Society.

[4]  Jan Wessnitzer,et al.  A Model of Non-elemental Associative Learning in the Mushroom Body Neuropil of the Insect Brain , 2007, ICANNGA.

[5]  Günther Palm,et al.  Comparison of Multiclass SVM Decomposition Schemes for Visual Object Recognition , 2005, DAGM-Symposium.

[6]  Martti Juhola,et al.  A scatter method for data and variable importance evaluation , 2012, Integr. Comput. Aided Eng..

[7]  Martti Juhola,et al.  Classification and retrieval on macroinvertebrate image databases , 2011, Comput. Biol. Medicine.

[8]  Nello Cristianini,et al.  Large Margin DAGs for Multiclass Classification , 1999, NIPS.

[9]  Thomas G. Dietterich,et al.  Automated insect identification through concatenated histograms of local appearance features: feature vector generation and region detection for deformable objects , 2007, 2007 IEEE Workshop on Applications of Computer Vision (WACV '07).

[10]  Moncef Gabbouj,et al.  Network of evolutionary binary classifiers for classification and retrieval in macroinvertebrate databases , 2010, 2010 IEEE International Conference on Image Processing.

[11]  Kristian Meissner,et al.  Statistical classification and proportion estimation - an application to a macroinvertebrate image database , 2010, 2010 IEEE International Workshop on Machine Learning for Signal Processing.

[12]  S. Kiranyaz,et al.  CLASSIFICATION AND RETRIEVAL ON MACROINVERTABRATE IMAGE DATABASES USING EVOLUTIONARY RBF NEURAL NETWORKS , 2011 .

[13]  Johan A. K. Suykens,et al.  Least Squares Support Vector Machine Classifiers , 1999, Neural Processing Letters.

[14]  Martti Juhola,et al.  Comparing the One-vs-One and One-vs-All Methods in Benthic Macroinvertebrate Image Classification , 2011, MLDM.

[15]  Ville Tirronen,et al.  Multiple Order Gradient Feature for Macro-Invertebrate Identification Using Support Vector Machines , 2009, ICANNGA.

[16]  Thomas M. Cover,et al.  Geometrical and Statistical Properties of Systems of Linear Inequalities with Applications in Pattern Recognition , 1965, IEEE Trans. Electron. Comput..

[17]  Martti Juhola,et al.  Kernel selection in multi-class support vector machines and its consequence to the number of ties in majority voting method , 2011, Artificial Intelligence Review.

[18]  Nello Cristianini,et al.  An Introduction to Support Vector Machines and Other Kernel-based Learning Methods , 2000 .