Directed Acyclic Graph Support Vector Machine Research Articles

A Directed Acyclic Graph Algorithm for Multi-class Classification Based on Twin Support Vector Machine

In order to deal with the inefficient and sample unbalance in multi-class classification by using support vector machine, a novel algorithm is proposed. The algorithm termed as Directed Acyclic Graph and Twin Support Vector Machine (DAG-TWSVM for short), is a combination of Directed Acyclic Graph Support Vector Machine (DAG-SVM) with those of Twin Support Vector Machine (TWSVM). The present DAG-TWSVM algorithm can not only get better classification accuracy, but also reduce the training time. The UCI datasets and Statlog datasets were used to test the novel algorithm, experiment results showed that DAG-TWSVM use less time in training than other multi-class SVMs, and outperformed other multi-class approaches on unbalanced dataset.

Directed acyclic graph support vector machines in classification of benthic macroinvertebrate samples

Freshwater areas are an evanescent resource in the globe and, hence, the water quality assessment and examination of human induced changes on aquatic ecosystems are important. Benthic macroinvertebrates are excellent indicators of the state of freshwater area due to their intermediate life cycle and their ability to react changes in an aquatic ecosystem. However, identifying the benthic macroinvertebrates is a slow and expensive task when the process is human-made. Thus, we have an excellent opportunity to save resources by using sophisticated machine learning and pattern recognition methods. In this research we applied Directed Acyclic Graph Support Vector Machine (DAGSVM) and six other baseline methods in automated benthic macroinvertebrate classification. We performed wide experimental tests with six different feature sets where one was chosen by Scatter method. Furthermore, in the case of DAGSVM we tested every feature set with seven kernel functions. Experimental results showed that Scatter method is a good feature selection algorithm and we obtained above 80 % accuracy in classifications which is a very good result for such a challenging classification task. From all tested classification methods DAGSVM showed to be the best alternative for this classification task.

Multicategory classification of 11 neuromuscular diseases based on microarray data using support vector machine.

We applied multicategory machine learning methods to classify 11 neuromuscular disease groups and one control group based on microarray data. To develop multicategory classification models with optimal parameters and features, we performed a systematic evaluation of three machine learning algorithms and four feature selection methods using three-fold cross validation and a grid search. This study included 114 subjects of 11 neuromuscular diseases and 31 subjects of a control group using microarray data with 22,283 probe sets from the National Center for Biotechnology Information (NCBI). We obtained an accuracy of 100%, relative classifier information (RCI) of 1.0, and a kappa index of 1.0 by applying the models of support vector machines one-versus-one (SVM-OVO), SVM one-versus-rest (OVR), and directed acyclic graph SVM (DAGSVM), using the ratio of genes between categories to within-category sums of squares (BW) feature selection method. Each of these three models selected only four features to categorize the 12 groups, resulting in a time-saving and cost-effective strategy for diagnosing neuromuscular diseases. In addition, a gene symbol, SPP1 was selected as the top-ranked gene by the BW method. We confirmed relationships between the gene (SPP1) and Duchenne muscular dystrophy (DMD) from a previous study. With our models as clinically helpful tools, neuromuscular diseases could be classified quickly using a computer, thereby giving a time-saving, cost-effective, and accurate diagnosis.

Electrocardiogram beat classification using empirical mode decomposition and multiclass directed acyclic graph support vector machine

In this paper, a classifier motivated from statistical learning theory, i.e., support vector machine, with a new approach based on multiclass directed acyclic graph has been proposed for classification of four types of electrocardiogram signals. The motivation for selecting Directed Acyclic Graph Support Vector Machine (DAGSVM) is to have more accurate classifier with less computational cost. Empirical mode decomposition and subsequently singular value decomposition have been used for computing the feature vector matrix. Further, fivefold cross-validation and particle swarm optimization have been used for optimal selection of SVM model parameters to improve the performance of DAGSVM. A comparison has been made between proposed algorithm and other two classifiers, i.e., K-Nearest Neighbor (KNN) and Artificial Neural Network (ANN). The DAGSVM has yielded an average accuracy of 98.96% against 95.83% and 96.66% for the KNN and the ANN, respectively. The results obtained clearly confirm the superiority of the DAGSVM approach over other classifiers.

Radar Emitter Signal Feature Extraction Based on Time-Frequency Atom Decomposition

In this paper, a novel approach based on time-frequency atom decomposition is presented to recognize the radar emitter signals. To decompose the signals into a linear expansion of time-frequency atoms, a fast matching pursuit (MP) algorithm, which is optimized by composite differential evolution (CoDE) algorithm, is introduced. The feature vectors of radar emitter signals are extracted based on the atoms generated in the process of decomposition. The Directed Acyclic Graph SVM (DAGSVM) is selected as the classifier to classify the feature vectors of different radar emitter signals.

GA Optimized Binary Tree SVM for Analog Circuit Fault Diagnosis

This paper presents a novel method of analog circuit fault diagnosis based on genetic algorithm (GA) optimized binary tree support vector machine (SVM). The real-valued coding genetic algorithm is used to optimize the binary tree structure. In optimization algorithm, we use roulette wheel selection operator, partially mapped crossover operator, inversion mutation operator. In simulation experiment, we use Monte-carlo analysis for 40kHz Sallen-Key bandpass filter and get transient response of ten faults. Then we extract feature vector by db3 wavelet packet transform and principal component analysis (PCA), and diagnose circuit faults by different SVM methods. Experiment results show the proposed method has the better classification accuracy than one-against-one (o-a-o), one-against-rest (o-a-r), Directed Acyclic Graph SVM (DAGSVM) and binary tree SVM (BT-SVM). It is suitable for practical use.

Classification of Fruits Using Computer Vision and a Multiclass Support Vector Machine

Automatic classification of fruits via computer vision is still a complicated task due to the various properties of numerous types of fruits. We propose a novel classification method based on a multi-class kernel support vector machine (kSVM) with the desirable goal of accurate and fast classification of fruits. First, fruit images were acquired by a digital camera, and then the background of each image was removed by a split-and-merge algorithm; Second, the color histogram, texture and shape features of each fruit image were extracted to compose a feature space; Third, principal component analysis (PCA) was used to reduce the dimensions of feature space; Finally, three kinds of multi-class SVMs were constructed, i.e., Winner-Takes-All SVM, Max-Wins-Voting SVM, and Directed Acyclic Graph SVM. Meanwhile, three kinds of kernels were chosen, i.e., linear kernel, Homogeneous Polynomial kernel, and Gaussian Radial Basis kernel; finally, the SVMs were trained using 5-fold stratified cross validation with the reduced feature vectors as input. The experimental results demonstrated that the Max-Wins-Voting SVM with Gaussian Radial Basis kernel achieves the best classification accuracy of 88.2%. For computation time, the Directed Acyclic Graph SVMs performs swiftest.

An automatic video content classification scheme based on combined visual features model with modified DAGSVM

Automatic video content classification attracts much attention from researchers in multimedia analysis because the management of video content is a challenging task. In this paper, a visual feature representation composed of editing, color, texture and motion features is proposed which is shown to be effective in differentiating among various video contents. A modified Directed Acyclic Graph Support Vector Machine (DAGSVM) model as the classifier is also presented. Experiments show that the features extracted have improved the discriminative ability between different video contents and the computational complexity has also been reduced. By introducing the DAG policy, the performance of the classifier has been enhanced and the classification results demonstrate the precision and effectiveness of this approach, compared with the other two classification methods. In addition, the proposed algorithm can be applied to video searching and harmful-video content filtering, etc.

Analyzing foreign exchange rates by rough set theory and directed acyclic graph support vector machines

Rough set theory (RST) and directed acyclic graph support vector machines (DAGSVM) are two emerging techniques in dealing with classification problems. The RST approach is able to select important features and generate rules from data. The SVM technique is powerful in solving classification problems with high generalization ability by applying the structure risk minimization principle. However, one particular model cannot capture all data patterns easily. This investigation presents a hybrid RST and DAGSVM model (RSTDAGSVM) to exploit the unique strengths of both RST and SVM in analyzing the movements of exchange rates. In the proposed hybrid model, the RST approach is used to extract the rules of exchange rate changes; and the DAGSVM technique is employed to deal with situations that cannot be included in the RST model. In addition, an immune algorithm and tabu search (IA/TS) method is applied to select parameters of SVM models. Experimental results reveal that the developed model achieves more accurate prediction results than either the RST model or the DAGSVM model on its own. Thus, the presented RSTDAGSVM model is a promising alternative for analyzing exchange rates.

A novel hierarchical speech emotion recognition method based on improved DDAGSVM

In order to improve the recognition accuracy of speech emotion recognition, in this paper, a novel hierarchical method based on improved Decision Directed Acyclic Graph SVM (improved DDAGSVM) is proposed for speech emotion recognition. The improved DDAGSVM is constructed according to the confusion degrees of emotion pairs. In addition, a geodesic distance-based testing algorithm is proposed for the improved DDAGSVM to give the test samples differently distinguished many decision chances. Informative features and SVM optimized parameters used in each node of the improved DDAGSVM are gotten by Genetic Algorithm (GA) synchronously. On the Chinese Speech Emotion Database (CSED) and the Audio-Video Emotion Database (AVED) recorded by our workgroup, the recognition experiment results reveal that, compared with multi-SVM, binary decision tree and traditional DDAGSVM, the improved DDAGSVM has the higher recognition accuracy with few selected informative features and moderate time for 7 emotions.

A NEW FAULT PATTERN RECOGNITION METHOD BASED ON WPT AND DAGSVM

A new pattern recognition method based on wavelet packet transform (WPT) and directed acyclic graph support vector machine (DAGSVM) is put forward for fault diagnosis of roller bearing. The fault pattern recognition model setup has two phases. The first phase is to extract the feature of faulty vibration signals from roller bearing by WPT via a db3 wavelet. The second phase is to use DAGSVM to recognize fault pattern of roller bearing. The testing results illustrates that WPT is more effective to diagnose fault types than the WT method. It is observed that among the strategy of multi-class SVM, DAGSVM acquires the highest accuracy, and therefore, this demonstrates the fact that suitable fault pattern recognition strategy can improve the overall performance of fault diagnosis. The present research illustrated that the features extracted by WPT represent the fault pattern of roller bearing, and the DAGSVM trained on these features achieved high recognition accuracies.

Solving large-scale multiclass learning problems via an efficient support vector classifier

Support vector machines (SVMs) are initially designed for binary classification. How to effectively extend them for multiclass classification is still an ongoing research topic. A multiclass classifier is constructed by combining SVM light algorithm with directed acyclic graph SVM (DAGSVM) method, named DAGSVM light. A new method is proposed to select the workings set which is identical to the working set selected by SVM light approach. Experimental results indicate DAGSVM light is competitive with DAGSMO. It is more suitable for practice use. It may be an especially useful tool for large-scale multiclass classification problems and lead to more widespread use of SVMs in the engineering community due to its good performance.

Bond rating using support vector machine

This paper deals with the application of support vector machine (SVM) for bond rating. The three commonly used methods for solving multi-class classification problems in SVM, one-against-all, one-against-one, and directed acyclic graph SVM (DAGSVM) are used. The performance of SVM is compared with several benchmarks. One real U.S. bond data is collected using the Fixed Investment Securities database (FISD) and the Compustat database. The experiment shows that SVM significantly outperforms the benchmarks. Among the three SVM based methods, there is the best performance in DAGSVM. Furthermore, an analysis of features shows that the generalization performance of SVM can be further improved by performing feature selection.

Binary tree of SVM: a new fast multiclass training and classification algorithm

We present a new architecture named Binary Tree of support vector machine (SVM), or BTS, in order to achieve high classification efficiency for multiclass problems. BTS and its enhanced version, c-BTS, decrease the number of binary classifiers to the greatest extent without increasing the complexity of the original problem. In the training phase, BTS has N - 1 binary classifiers in the best situation (N is the number of classes), while it has log4/3 ((N + 3)/4) binary tests on average when making a decision. At the same time the upper bound of convergence complexity is determined. The experiments in this paper indicate that maintaining comparable accuracy, BTS is much faster to be trained than other methods. Especially in classification, due to its Log complexity, it is much faster than directed acyclic graph SVM (DAGSVM) and ECOC in problems that have big class number.

A method to make multiple hypotheses with high cumulative recognition rate using SVMs

This paper describes a method to make multiple hypotheses with high cumulative recognition rate using SVMs. To make just a single hypothesis by using SVMs, it has been shown that Directed Acyclic Graph Support Vector Machines (DAGSVM) is very good with respect to recognition rate, learning time and evaluation time. However, DAGSVM is not directly applicable to make multiple hypotheses. In this paper, we propose a hybrid method of DAGSVM and Max-Win algorithm. Based on the result of DAGSVM, a limited set of classes are extracted. Then, Max-Win algorithm is applied to the set. We also provide the experimental results to show that the cumulative recognition rate of our method is as good as the Max-Win algorithm, and that the execution time is almost as fast as DAGSVM.

A comparison of methods for multiclass support vector machines

Support vector machines (SVMs) were originally designed for binary classification. How to effectively extend it for multiclass classification is still an ongoing research issue. Several methods have been proposed where typically we construct a multiclass classifier by combining several binary classifiers. Some authors also proposed methods that consider all classes at once. As it is computationally more expensive to solve multiclass problems, comparisons of these methods using large-scale problems have not been seriously conducted. Especially for methods solving multiclass SVM in one step, a much larger optimization problem is required so up to now experiments are limited to small data sets. In this paper we give decomposition implementations for two such "all-together" methods. We then compare their performance with three methods based on binary classifications: "one-against-all," "one-against-one," and directed acyclic graph SVM (DAGSVM). Our experiments indicate that the "one-against-one" and DAG methods are more suitable for practical use than the other methods. Results also show that for large problems methods by considering all data at once in general need fewer support vectors.