Local Support Vector Machine Research Articles

New hierarchical approach for microaneurysms detection with matched filter and machine learning.

Microaneurysms are regarded as the first signs of diabetic retinopathy (DR), but the microaneurysms are not clear in the color retinal images, and many researches were studied to detect and locate these lesions. In this paper, a new hierarchical computing-aided diagnosis approach is proposed for the microaneurysms detection by using the multi-scale and multi-orientation sum of matched filter (MMMF) and machine learning, where 37 dimensional features are extracted from each candidate. Furthermore, several classifiers such as the k-nearest neighbor (kNN), local linear discrimination analysis (LLDA) and support vector machine (SVM) are modified to distinguish the true microaneurysms from the false ones, which is a typical unbalanced classification problem. The effectiveness of the proposed method is verified through the training set of a publicly available database, and the experiment results show that the proposed method has better detection performance including the receiver operating characteristic (ROC) curve and the free-response receiver operating characteristic (FROC) curve. Moreover, the proposed method with 37 dimensional features outperforms that with other features and has a sensitivity from 1/8 to 8 with the average of all seven points being 0.286 tested on the same database.

Detection of masses in mammograms with adaption to breast density using genetic algorithm, phylogenetic trees, LBP and SVM

Breast cancer is the second commonest type of cancer in the world, and the commonest among women, corresponding to 22% of the new cases every year. This work presents a new computational methodology, which helps the specialists in the detection of breast masses based on the breast density. The proposed methodology is divided into stages with the objective of overcoming several difficulties associated with the detection of masses. In many of these stages, we brought contributions to the areas. The first stage is intended to detect the type of density of the breast, which can be either dense or non-dense. We proposed an adaptive algorithm capable of analyzing and image and telling if it is dense or non-dense. The first stage consists in the segmentation of the regions that look like masses. We propose a novel use of the micro-genetic algorithm to create a texture proximity mask and select the regions suspect of containing lesions. The next stage is the reduction of false positives, which were generated in the previous stage. To this end, we proposed two new approaches. The first reduction of false positives used DBSCAN and a proximity ranking of the textures extracted from the ROIs. In the second reduction of false positives, the resulting regions have their textures analyzed by the combination of Phylogenetic Trees, Local Binary Patterns and Support Vector Machines (SVM). A micro-genetic algorithm was used to choose the suspect regions that generate the best training models and maximize the classification of masses and non-masses used in the SVM. The best result produced a sensitivity of 92.99%, a rate of 0.15 false positives per image and an area under the FROC curve of 0.96 in the analysis of the non-dense breasts; and a sensitivity of 83.70%, a rate of 0.19 false positives per image and an area under the FROC curve of 0.85, in the analysis of the dense breasts.

Comparative Study on Biometric Iris Recognition based on Hamming Distance and Multi Block Local Binary Pattern

Personal identification based on biometric is most essential to ensure security. Recognition based on iris unique texture is a reliable, simple and fast. The abundant as well as unique patterns of iris are extracted. Matrix format template is generated that contains 4800 elements for each iris. Multi block local binary pattern, hamming distance and support vector machine performs matching based on the template’s unique features of iris. The experimental results of this proposed work illustrate a better performance. The popular CASIA (Chinese Academy of Sciences – Institute of Automation) iris database with hundred users’ eye image samples are experimented to prove, that the multi block local binary pattern is comparatively better with minimal true rejection rate.

Small Time-Scale Network Tidal Water Level Forecasting Based on Quantum Neural Network

Zhao, K.; Wang, H., and Guo, T., 2015. Small time-scale network traffic prediction on the basis of Quantum Neural Network.This paper proposed a quantum neural network model consist of quantum bit, universal quantum gates and quantum weighted, gave the learning algorithm based on improved PRP conjugate gradient, in order to improve the convergence speed of the network and the network performance, and further proved the global convergence of the algorithm in theory. Compared with the existing localized support vector machine (SVM) regression and flexible neural tree model, the QNN model proposed in this paper have a higher prediction precision and low computational complexity for small scale tidal water level prediction, its convergence speed and robustness are better than BP network and quantum weighted neural network.

Assessment of the EuroSCORE risk scoring system for patients undergoing coronary artery bypass graft surgery in a group of Iranian patients.

Background and Aims:Previous studies around the world indicated validity and accuracy of European System for Cardiac Operative Risk Evaluation (EuroSCORE) risk scoring system we evaluated the EuroSCORE risk scoring system for patients undergoing coronary artery bypass graft (CABG) surgery in a group of Iranian patients.Materials and Methods:In this cohort 2220 patients more than 18 years, who were performed CABG surgery in Massih Daneshvari Hospital, from January 2004 to March 2010 were recruited. Predicted mortality risk scores were calculated using logistic EuroSCORE and Acute Physiology and Chronic Health Evaluation II (APACHE II) and compared with observed mortality. Calibration was measured by the Hosmer–Lemeshow (HL) test and discrimination by using the receiver operating characteristic (ROC) curve area.Results:Of the 2220 patients, in hospital deaths occurred in 270 patients (mortality rate of 12.2%). The accuracy of mortality prediction in the logistic EuroSCORE and APACHE II model was 89.1%; in the local EuroSCORE (logistic) was 91.89%; and in the local EuroSCORE support vector machines (SVM) was 98.6%. The area under curve for ROC curve, was 0.724 (95% confidence interval [CI]: 0.57–0.88) for logistic EuroSCORE; 0.836 (95% CI: 0.731–0.942) for local EuroSCORE (logistic); 0.978 (95% CI: 0.937–1) for Local EuroSCORE (SVM); and 0.832 (95% CI: 0.723–0.941) for APACHE II model. The HL test showed good calibration for the local EuroSCORE (SVM), APACHE II model and local EuroSCORE (logistic) (P = 0.823, P = 0.748 and P = 0.06 respectively); but there was a significant difference between expected and observed mortality according to EuroSCORE model (P = 0.033).Conclusion:We detected logistic EuroSCORE risk model is not applicable on Iranian patients undergoing CABG surgery.

Identification of Mitral Annulus Hinge Point Based on Local Context Feature and Additive SVM Classifier.

The position of the hinge point of mitral annulus (MA) is important for segmentation, modeling and multimodalities registration of cardiac structures. The main difficulties in identifying the hinge point of MA are the inherent noisy, low resolution of echocardiography, and so on. This work aims to automatically detect the hinge point of MA by combining local context feature with additive support vector machines (SVM) classifier. The innovations are as follows: (1) designing a local context feature for MA in cardiac ultrasound image; (2) applying the additive kernel SVM classifier to identify the candidates of the hinge point of MA; (3) designing a weighted density field of candidates which represents the blocks of candidates; and (4) estimating an adaptive threshold on the weighted density field to get the position of the hinge point of MA and exclude the error from SVM classifier. The proposed algorithm is tested on echocardiographic four-chamber image sequence of 10 pediatric patients. Compared with the manual selected hinge points of MA which are selected by professional doctors, the mean error is in 0.96 ± 1.04 mm. Additive SVM classifier can fast and accurately identify the MA hinge point.

A Novel Multi-Feature Descriptor for Human Detection Using Cascaded Classifiers in Static Images

Combining multiple kinds of features is useful to achieve the state of the art performance for human detection. But combining more features will result in high dimensional feature descriptors, which is time-consuming for feature extraction and detection. How to exploit different kinds of features and reduce the dimension of feature descriptor are challenging problems. A novel multi-feature descriptor (MFD) combining Optimal Histograms of Oriented Gradients (OHOG), Local Binary Patterns (LBP) and Color Self-Similarity in Neighbor (NCSS) is proposed. Firstly, a discriminative feature selection and combination strategy is introduced to obtain distinctive local HOGs and construct OHOG feature. OHOG combines local discriminative and correlated information, which improves the classification performance compared with HOG. Besides, LBP describes texture feature of human appearance. Finally, a compact and lower dimensional feature NCSS is proposed to encode the self-similarity of color histograms in limited neighbor sub-regions instead of global regions. The proposed MFD describes human appearance from gradient, texture and color features, which can complement each other and improve the robustness of human description. To further improve detection speed without decreasing accuracy, we cascade early stages of Adaboost based on selected local HOGs and SVM classifier based on MFD. The former part can reject most non-human detection windows quickly and the final SVM classifier can guarantee a high accuracy. Experimental results on public dataset show that the proposed MFD and cascaded classifiers framework can achieve promising results both in accuracy and detection speed.

Face Gender Recognition Research Based on Local Features and Support Vector Machine

In this paper, we proposed a face gender recognition method based on local features and SVM. First, we divide the face image into five parts which are used to instead of the whole face for better recognition performance. Second, we use CS to extract local features of these five parts. Then, we respectively train five single SVM classifiers to achieve one to one feature recognition for local features. Finally, decision information fusion is used to achieve the final classification. Because SVM were successfully used to solve numerous pattern recognition problems and is mainly used to solve two-classification problem, selecting SVM to do gender recognition in our method has the obvious superiority. After a lot of experiments, results show that the proposed method in this paper is stable and effective, greatly improving the efficiency of face gender recognition.

Feature-Weighted Local Support Vector Machine of Particle Swarm Optimization

Local support vector machine gives the feature same weight in classification. In fact, many datasets have some weak or irrelevant features related to the classification. Thus giving features same weight may reduce the classification accuracy of local support vector machine.This paper puts forward a new local support vector machine that the feature weight is optimized by PSO (Particle Swarm Optimization), it is tested on the international standard UCI data sets and the images of tree taxonomy data sets, the results show that the accuracy of the algorithm we proposed is better than the general local support vector machine.

Detecting China’s Urban Expansion Over the Past Three Decades Using Nighttime Light Data

China has experienced a rapid urban expansion over the past three decades because of its accelerated economic growth. In this study, we detected and analyzed the urban expansion of China during this period using multi-temporal Defense Meteorological Satellite Program Operational Linescan System (DMSP-OLS) nighttime light data and multi-source Normalized Difference Vegetation Index (NDVI) data. First, an intercalibration was performed to improve the continuity and comparability of the nighttime light data from 1992 to 2010. The nighttime light and NDVI data were then subjected to a local support vector machine (SVM) based region-growing method to extract the urban areas from 1992 to 2010. The urban areas from 1981 to 1991 were identified using the areas in 1992 and NDVI data, based on the hypothesis that China’s urban expansion continued during this period. Finally, the extracted time-series urban maps were validated with Landsat images. The proposed local SVM-based region-growing method performed better than a local thresholding method and a global SVM-based region-growing method according to visual and quantitative comparisons of the urban boundaries and areas. We also analyzed the expansion rates to understand the dynamics of the urban areas in China and in its seven economic regions. In particular, the urban expansion patterns were investigated in three typical urban agglomerations, i.e., Beijing–Tianjin–Hebei, Yangtze River Delta, and Pearl River Delta. The proposed urban expansion direction, urban expansion intensity, and relative ratio of urban expansion demonstrated the regional variation among the three urban agglomerations.

FACIAL EXPRESSION RECOGNITION BASED ON LOCAL BINARY PATTERN FEATURES AND SUPPORT VECTOR MACHINE

Facial expression recognition has been researched much in recent years because of their applications in intelligent communication systems. Many methods have been developed based on extracting Local Binary Pattern (LBP) features associating different classifying techniques in order to get more and more better effects of facial expression recognition. In this work, we propose a novel method for recognizing facial expressions based on Local Binary Pattern features and Support Vector Machine with two effective improvements. First is the preprocessing step and second is the method of dividing face images into nonoverlap square regions for extracting LBP features. The method was experimented on three typical kinds of database: small (213 images), medium (2040 images) and large (5130 images). Experimental results show the effectiveness of our method for obtaining remarkably better recognition rate in comparison with other methods.

Realistic human action recognition by Fast HOG3D and self-organization feature map

Nowadays, local features are very popular in vision-based human action recognition, especially in "wild" or unconstrained videos. This paper proposes a novel framework that combines Fast HOG3D and self-organization feature map (SOM) network for action recognition from unconstrained videos, bypassing the demanding preprocessing such as human detection, tracking or contour extraction. The contributions of our work not only lie in creating a more compact and computational effective local feature descriptor than original HOG3D, but also lie in first successfully applying SOM to realistic action recognition task and studying its training parameters' influence. We mainly test our approach on the UCF-YouTube dataset with 11 realistic sport actions, achieving promising results that outperform local feature-based support vector machine and are comparable with bag-of-words. Experiments are also carried out on KTH and UT-Interaction datasets for comparison. Results on all the three datasets confirm that our work has comparable, if not better, performance comparing with state-of-the-art.

Targeted Local Support Vector Machine for Age-Dependent Classification

We develop methods to accurately predict whether presymptomatic individuals are at risk of a disease based on their various marker profiles, which offers an opportunity for early intervention well before definitive clinical diagnosis. For many diseases, existing clinical literature may suggest the risk of disease varies with some markers of biological and etiological importance, for example, age. To identify effective prediction rules using nonparametric decision functions, standard statistical learning approaches treat markers with clear biological importance (e.g., age) and other markers without prior knowledge on disease etiology interchangeably as input variables. Therefore, these approaches may be inadequate in singling out and preserving the effects from the biologically important variables, especially in the presence of potential noise markers. Using age as an example of a salient marker to receive special care in the analysis, we propose a local smoothing large margin classifier implemented with support vector machine (SVM) to construct effective age-dependent classification rules. The method adaptively adjusts age effect and separately tunes age and other markers to achieve optimal performance. We derive the asymptotic risk bound of the local smoothing SVM and perform extensive simulation studies to compare with standard approaches. We apply the proposed method to two studies of premanifest Huntington’s disease (HD) subjects and controls to construct age-sensitive predictive scores for the risk of HD and risk of receiving HD diagnosis during the study period. Supplementary materials for this article are available online.

Natural gas pipeline small leakage feature extraction and recognition based on LMD envelope spectrum entropy and SVM

As it is difficult to identify the scale and aperture of small leaks occurring in a natural gas pipeline, this paper proposes a small leak feature extraction and recognition method based on local mean decomposition (LMD) envelope spectrum entropy and support vector machine (SVM). First, LMD is used to decompose the leakage signals into several FM–AM signals, i.e. into product function (PF) components. Then, based on their kurtosis features, the principal PF components that contain most of the leakage information are selected. Wavelet packet decomposition and energy methods are used to analyze and then reconstruct the principal PF components. The Hilbert transform is applied to these reconstructed principal PF components in order to acquire the envelope spectrum, from which the envelope spectrum entropy is obtained. Finally the normalized envelope spectrum entropy features are input into the SVM as leakage feature vectors in order to enable leak aperture category identification. By analyzing the acquired pipeline leakage signals in field experiments, it shows that this method can effectively identify different leak categories.

Fly Wing Biometrics Using Modified Local Binary Pattern, SVMs and Random Forest

This paper presents an efficient approach for classification of the gender of a common fruit fly, Drosophila melanogaster, based on their wing's texture. The novelty of this research effort is that a Modified Local Binary Pattern (MLBP), which combines both the sign and magnitude features for the improvement of fly wing's texture classification performance, is applied. The extracted features are then used to classify the gender of the fruit fly by using the Support Vector Machines (SVMs) and Random Forest (RF). We validate the performance of the proposed scheme on two fly wing datasets. The highest accuracy achieved by the proposed approach is 94%. In this paper, we limit our approach to gender classification; however, this effort can be extended to explore important characteristics of a fly using wing's texture analysis.

A quantitative estimation technique for welding quality using local mean decomposition and support vector machine

The experimental nonlinear time series of welding current contain the arc feature information related to welding quality. The local mean decomposition (LMD) combining with the support vector machine (SVM) is put forward to quantitatively estimate the rationality of welding parameters and welding formation quality. The LMD is used to investigate the time---frequency distribution of arc energy, and the energy entropy is employed to quantitatively judge the welding arc characteristics related to welding quality. The collected current signal is decomposed into a number of product functions (PFs) by LMD. The energy entropy of each PF is calculated to establish the welding arc energy feature vectors, which are inputted into support vector machine classifier. The LMD combining with SVM can quantitatively estimate the time---frequency energy distribution characteristics of the arc current signal at different welding parameters and welding formation quality. Experimental results are provided to confirm the effectiveness of this approach to estimate the rationality of welding parameters and welding formation quality.

Gas Pipeline Small Leak Aperture Classification Based on Local Mean Decomposition Envelope Spectrum Entropy and SVM

When small leak occurs in the natural gas pipeline, it is difficult to identify the leak scale and aperture. It is proposed a small leak aperture recognition method based on local mean decomposition(LMD) envelope spectrum entropy and SVM. The leakage signals are decomposed into a number of production functions(PFs) components which have physical significance instantaneous frequencies. And then calculate the PFs kurtosis values and according to this select the principal PF components which contain most of leakage information. Further the wavelet packet decomposition and band energy distribution method are used to analyze the principal PF components and then reconstruct them. The Hilbert transform is applied to these reconstructed principal PF components and the corresponding envelope spectrums are obtained. Combining the concept of information entropy, the envelope spectrum entropy is proposed and calculates the entropy values. The normalized envelope spectrum entropy as the leakage feature is input the support vector machine(SVM) and the leak aperture classification is accomplished. By analyzing the acquired pipeline leakage signals in the field experiments, the results show that this method can effectively identify the different leak apertures.

A local least square support vector machine prediction algorithm of small scale network traffic based on correlation analysis

Real-time monitoring and forecasting technology for network traffic has played an important role in network management. Effective network traffic prediction could analyze and solve problems before overload occurs, which significantly improves network availability. In this paper, after the vulnerability of traditional nonlinear prediction method in forecasting modeling is analyzed, the relevant local (RL) forecast which is based on correlation analysis and the parameter optimization method based on pattern search (PS) is introduced. Using the correlation analysis, the optimal training subset is chosen from time-and distance-correlated training samples. On this basis, the prediction model is established by LSSVM. Finally network traffic dataset collected from wired campus networks is studied for our experiments. And the results show that the relevant local LSSVM prediction method whose training set and parameters have been automatically optimized can effectively predict the small scale traffic measurement data, and RL-LSSVM traffic forecasting algorithm exhibits significantly good prediction accuracy for the data set compared with previous algorithm.

Aero-Engine Fault Diagnosis Using Improved Local Discriminant Bases and Support Vector Machine

This paper presents an effective approach for aero-engine fault diagnosis with focus on rub-impact, through combination of improved local discriminant bases (LDB) with support vector machine (SVM). The improved LDB algorithm, using both the normalized energy difference and the relative entropy as quantification measures, is applied to choose the optimal set of orthogonal subspaces for wavelet packet transform- (WPT-) based signal decomposition. Then two optimal sets of orthogonal subspaces have been obtained and the energy features extracted from those subspaces appearing in both sets will be selected as input to a SVM classifier to diagnose aero-engine faults. Experiment studies conducted on an aero-engine rub-impact test system have verified the effectiveness of the proposed approach for classifying working conditions of aero-engines.

Targeted Local Support Vector Machine for Age-Dependent Classification

Targeted Local Support Vector Machine for Age-Dependent Classification