Performance Of Support Vector Machine Research Articles

Support Vector Machines for the Estimation of Specific Charge in Tunnel Blasting

Mine tunnels, short transportation tunnels, and hydro-power plan underground spaces excavations are carried out based on Drilling and Blasting (D&B) method. Determination of specific charge in tunnel D&B, according to the involved parameters, is very significant to present an appropriate D&B design. Suitable explosive charge selection and distribution lead to reduced undesirable effects of D&B such as inappropriate pull rate, over-break, under-break, unauthorized ground vibration, air blast, and fly rock. So far, different models are presented to estimate specific charge in tunnel blasting. In this study, 332 data sets, including geomechanical characteristics, D&B, and specific charge are gathered from 33 tunnels. The data are related to three dams and hydropower plans in Iran (Gotvand, Masjed-Solayman, and Siah-Bishe). Specific charge is modeled in inclined hole cut drilling pattern. In this regard, Support Vector Machine (SVM) algorithm based on polynomial Kernel function is used as a tool for modeling. Rock Quality Designation (RQD) index, Uniaxial Compressive Strength (UCS), tunnel cross-section area, maximum depth of blast hole, and blast hole coupling ratio are considered as independent input variables and the specific charge is considered as a dependent output variable. The modeling results confirm the acceptable performance of SVM in specific charge estimation with minimum error.

Improved non-contact variable-frequency AC impedance instrument for cement hydration and pore structure based on SVM calibration method

The non-contact alternating current (AC) impedance method is widely used in the study of electrochemical properties of cement-based materials,the instrument of this paper provide a new way for the study of pore structure of cement-based materials. Due to the loss of magnetic materials and the parasitic capacitance of measuring circuit, the ratio difference and phase difference of current sensor inevitably exit. Traditionally the standard resistor calibration method is used to determine the ratio error and phase error under different frequencies based on polynomial fitting (FT). In this paper, a machine-learning support vector machine (SVM) model is used to predict the calibration data with a small test sample. For the accuracy assessment, SVM model has the best comprehensive performance among FT, artificial neural network (ANN). The actual impedance test of typical cement-based materials under different frequencies is given and the results verify the excellent performance of SVM.

Comparative Analysis of Image Classification Algorithms for Face Mask Detection

Background: The COVID-19 pandemic remains a problem in 2021. Health protocols are needed to prevent the spread, including wearing a face mask. Enforcing people to wear face masks is tiring. AI can be used to classify images for face mask detection. There are a lot of image classification algorithm for face mask detection, but there are still no studies that compare their performance.Objective: This study aims to compare the classification algorithms of classical machine learning. They are k-nearest neighbors (KNN), support vector machine (SVM), and a widely used deep learning algorithm for image classification which is convolutional neural network (CNN) for face masks detection.Methods: This study uses 5 and 3 cross-validation for assessing the performance of KNN, SVM, and CNN in face mask detection.Results: CNN has the best average performance with the accuracy of 0.9683 and average execution time of 2,507.802 seconds for classifying 3,725 faces with mask and 3,828 faces without mask images.Conclusion: For a large amount of image data, KNN and SVM can be used as temporary algorithms in face mask detection due to their faster execution times. At the same time, CNN can be trained to form a classification model. In this case, it is advisable to use CNN for classification because it has better performance than KNN and SVM. In the future, the classification model can be implemented for automatic alert system to detect and warn people who are not wearing face masks.

Kernel Based Data-Adaptive Support Vector Machines for Multi-Class Classification

Imbalanced data exist in many classification problems. The classification of imbalanced data has remarkable challenges in machine learning. The support vector machine (SVM) and its variants are popularly used in machine learning among different classifiers thanks to their flexibility and interpretability. However, the performance of SVMs is impacted when the data are imbalanced, which is a typical data structure in the multi-category classification problem. In this paper, we employ the data-adaptive SVM with scaled kernel functions to classify instances for a multi-class population. We propose a multi-class data-dependent kernel function for the SVM by considering class imbalance and the spatial association among instances so that the classification accuracy is enhanced. Simulation studies demonstrate the superb performance of the proposed method, and a real multi-class prostate cancer image dataset is employed as an illustration. Not only does the proposed method outperform the competitor methods in terms of the commonly used accuracy measures such as the F-score and G-means, but also successfully detects more than 60% of instances from the rare class in the real data, while the competitors can only detect less than 20% of the rare class instances. The proposed method will benefit other scientific research fields, such as multiple region boundary detection.

Deep Neural Network to Differentiate Brain Activity Between Patients With First-Episode Schizophrenia and Healthy Individuals: A Multi-Channel Near Infrared Spectroscopy Study

Backgrounds: Reduced brain cortical activity over the frontotemporal regions measured by near infrared spectroscopy (NIRS) has been reported in patients with first-episode schizophrenia (FES). This study aimed to differentiate between patients with FES and healthy controls (HCs) on basis of the frontotemporal activity measured by NIRS with a support vector machine (SVM) and deep neural network (DNN) classifier. In addition, we compared the accuracy of performance of SVM and DNN.Methods: In total, 33 FES patients and 34 HCs were recruited. Their brain cortical activities were measured using NIRS while performing letter and category versions of verbal fluency tests (VFTs). The integral and centroid values of brain cortical activity in the bilateral frontotemporal regions during the VFTs were selected as features in SVM and DNN classifier.Results: Compared to HCs, FES patients displayed reduced brain cortical activity over the bilateral frontotemporal regions during both types of VFTs. Regarding the classifier performance, SVM reached an accuracy of 68.6%, sensitivity of 70.1%, and specificity of 64.6%, while DNN reached an accuracy of 79.7%, sensitivity of 88.8%, and specificity of 74.9% in the classification of FES patients and HCs.Conclusions: Compared to findings of previous structural neuroimaging studies, we found that using DNN to measure the NIRS signals during the VFTs to differentiate between FES patients and HCs could achieve a higher accuracy, indicating that NIRS can be used as a potential marker to classify FES patients from HCs. Future additional independent datasets are needed to confirm the validity of our model.

Improvement of multi‐layer soil moisture prediction using support vector machines and ensemble Kalman filter coupled with remote sensing soil moisture datasets over an agriculture dominant basin in China

AbstractSoil moisture prediction is of great importance in crop yield forecasting and drought monitoring. In this study, the multi‐layer root zone soil moisture (0‐5, 0‐10, 10‐40 and 40‐100 cm) prediction is conducted over an agriculture dominant basin, namely the Xiang River Basin, in southern China. The support vector machines (SVM) coupled with dual ensemble Kalman filter (EnKF) technique (SVM‐EnKF) is compared with SVM for its potential capability to improve the efficiency of soil moisture prediction. Three remote sensing soil moisture products, namely SMAP, ASCAT and AMSR2, are evaluated for their performance in multi‐layer soil moisture prediction with SVM and SVM‐EnKF, respectively. Multiple cases are designed to investigate the performance of SVM, the effectiveness of coupling dual EnKF technique and the applicability of the remote sensing products in soil moisture prediction. The main results are as follows: (a) The efficiency of soil moisture prediction with SVM using meteorological variables as inputs is satisfactory for the surface layers (0‐5 and 0‐10 cm), while poor for the root zone layers (10‐40 and 40‐100 cm). Adding SMAP as input to SVM can improve its performance in soil moisture prediction, with more than 47% increase in theR‐value and at least 11% reduction in RMSE for all layers. However, adding ASCAT or AMSR2 has no improvement for its performance. (b) Coupling dual EnKF can significantly improve the performance of SVM in the soil moisture prediction of both surface and the root zone layers. The increase inR‐value is above 80%, while the reduction in BIAS and RMSE is respectively more than 90% and 63%. However, adding remote sensing soil moisture products as inputs can no further improve the performance of SVM‐EnKF. (c) The SVM‐EnKF can eliminate the influence of remote sensing soil moisture extreme values in soil moisture prediction, therefore, improve its accuracy.

An improved cardiac arrhythmia classification using an RR interval-based approach

Accurate and early detection of cardiac arrhythmia present in an electrocardiogram (ECG) can prevent many premature deaths. Cardiac arrhythmia arises due to the improper conduction of electrical impulses throughout the heart. In this paper, we propose an improved RR interval-based cardiac arrhythmia classification approach. The Discrete Wavelet Transform (DWT) and median filters were used to remove high-frequency noise and baseline wander from the raw ECG. Next, the processed ECG was segmented after the determination of the QRS region. We extracted the primary feature RR interval and other statistical features from the beats to classify the Normal, Premature Ventricular Contraction (PVC), and Premature Atrial Contraction (PAC). The K-Nearest Neighbour (k-NN), Support Vector Machine (SVM), Decision Tree (DT), Naïve Bayes (NB), and Random Forest (RF) classifier were utilised for classification. Overall performance of SVM with Gaussian kernel achieved Se % = 99.28, Sp % = 99.63, +P % = 99.28, and Acc % = 99.51, which is better than the other classifiers used in this method. The obtained results of the proposed method are significantly better and more accurate.

Evaluation of SVM performance in the detection of lung cancer in marked CT scan dataset

<p>This paper concerns the development/analysis of the IQ-OTH/NCCD lung cancer dataset. This CT-scan dataset includes more than 1100 images of diagnosed healthy and tumorous chest scans collected in two Iraqi hospitals. A computer system is proposed for detecting lung cancer in the dataset by using image-processing/computer-vision techniques. This includes three preprocessing stages: image enhancement, image segmentation, and feature extraction techniques. Then, support vector machine (SVM) is used at the final stage as a classification technique for identifying the cases on the slides as one of three classes: normal, benign, or malignant. Different SVM kernels and feature extraction techniques are evaluated. The best accuracy achieved by applying this procedure on the new dataset was 89.8876%.</p>

Default Prediction of Automobile Credit Based on Support Vector Machine

Automobile credit business has developed rapidly in recent years, and corresponding default phenomena occur frequently. Credit default will bring great losses to automobile financial institutions. Therefore, the successful prediction of automobile credit default is of great significance. Firstly, the missing values are deleted, then the random forest is used for feature selection, and then the sample data are randomly grouped. Finally, six prediction models of support vector machine (SVM), random forest and k-nearest neighbor (KNN), logistic, decision tree, and artificial neural network (ANN) are constructed. The results show that these six machine learning models can be used to predict the default of automobile credit. Among these six models, the accuracy of decision tree is 0.79, which is the highest, but the comprehensive performance of SVM is the best. And random grouping can improve the efficiency of model operation to a certain extent, especially SVM.

Prognosis Prediction of Stroke based on Machine Learning and Explanation Model

The prognosis prediction of stroke is of great significance to its prevention and treatment. This paper used machine learning to predict stroke prognosis, and use SHAP method to make feature importance and single sample analysis. Firstly, feature engineering, use Borderline-SMOTE algorithm to deal with data imbalance, use Support Vector Machine(SVM) to build a prognostic prediction model, and use Random Forest(RF), Decision Tree(DT), Logistic Regression(LR) for comparative analysis, and find the performance of SVM after feature engineering better than other models, the accuracy, specificity, F1 score, AUC value reach 0.8306, 0.8356, 0.8415 and 0.9140. Then, the model was further analyzed for explainability, and it was found that the top three causes of the disease were Glasgow Coma Score, NIHSS and atrial fibrillation. Finally, try to analysis a single sample, which is performed to determine that the patient is a low-risk patient, and suffering from atrial fibrillation is the largest potential risk factor for the patient.

Feature optimization using Backward Elimination and Support Vector Machines (SVM) algorithm for diabetes classification

Diabetes is a disease that occurs when the blood glucose level is higher than normal and also leads to health problems. Early and accurate diagnosis needs to be carried out on individuals affected by this disease. Furthermore, excellent treatment needs to be provided to prevent worse situations. Some studies have used several machine learning methods to diagnose diabetes. Furthermore, in this study, the Backward Elimination and Support Vector Machine (SVM) algorithm was used to classify the PIMA Indians diabetes dataset. It consisted of 268 diabetic and 500 non-diabetic patients with eight attributes. Backward Elimination is a feature selection method used to remove irrelevant features based on the linear regression model. Using this method, the right features for the model was expected. This method has some advantages which include increasing training time, decreasing complexity and improving performance and accuracy. Therefore, the performance of SVM improved. Based on the experiments, it was discovered that by combining feature selection algorithm (backward elimination) and SVM, the highest accuracy obtained was 85.71% using 90% data training. Therefore, it was concluded that Backward Elimination combined with SVM algorithm is an excellent method to classify diabetes by using the PIMA Indians diabetes dataset.

Feature‐level fusion of laser‐induced breakdown spectroscopy and Raman spectroscopy for improving support vector machine in clinical bacteria identification

AbstractIn clinical field, the diagnosis of many diseases and their development stages depend on the detection of the corresponding bacteria. Raman spectroscopy and laser‐induced breakdown spectroscopy (LIBS) are two novel spectral diagnostic technologies for clinical bacteria identification. Both of them have been used in clinical detection combined with optimized support vector machine (SVM). In this paper, two feature‐level fusion methods (before feature selection fusion [BFSF] and after feature selection fusion [AFSF]) were proposed to improve the performance of SVM classifier and reduce the analyzing time (including the parameter tuning, model training, and testing time) simultaneously by combining data of LIBS and Raman spectroscopy. Using the most important 10 feature lines as the inputs of the optimized classifier, the analyzing time could be reduced to 1 to 2 min and the correct classification rate (CCR) achieved 95.67%. Without optimizing SVM parameters, the two proposed methods could achieve rapid and accurate classification of pathogenic bacteria further. The AFSF method showed better results with less fusion features while the BFSF method achieved higher CCR at 100% with more features. Both two methods costed around 0.2 s for analysis. These indicate that the proposed feature‐level fusion methods can improve the performance of SVM for bacteria detection. Maintaining the highest diagnostic accuracy, they can achieve the minimum analyzing time.

Comparison of the performances of GEP, ANFIS, and SVM artifical intelligence models in rainfall simulaton

In this paper, evaluation the performances of GEP (gene expression programming), ANFIS (adaptive fuzzy interference system), and SVM (support vector machine) artificial intelligence models in two scales of daily and monthly rainfall data from Urmia meteorological station (Iran) and monthly rainfall data from Diata meteorological station (India) was used in rainfall simulation. The correlation coefficient of observed and simulated values was evaluated by the R2 criterion, simulation error was evaluated by the root mean square error (RMSE), and MB criteria and model efficiency were evaluated by the Nash-Sutcliffe method. The results show that the correlation coefficients in the GEP model based on daily data from Urmia station and monthly data from Diata station are 23 and 58%, respectively, and R2 in simulation with GEP is estimated to be 55% lower than with the other two models. The R2 range in both ANFIS and SVM models varies from 91 to 93%. On average, the RMSE values in the GEP simulation are 50% and 55% higher than the ANFIS ratio for daily and monthly data at the two stations, respectively, and the RMSE values of ANFIS model are 1% and 3% higher than those of the SVM at Urmia and Diata stations, respectively. The bias values of the GEP model are 72% and 60% higher than those of ANFIS at Urmia and Diata stations, respectively. The GEP efficiency factors are 56% and 61% lower than those of ANFIS at Urmia and Diata stations, respectively. And the ANFIS efficiency ratio is 1 and 2% lower than SVM in Urmia and Diata stations, respectively. Therefore, rainfall simulation with the SVM model is associated with a lower error rate and better efficiency, the ANFIS model is close to the efficiency of SVM, and the GEP model is not suitable for rainfall simulation.

Cross-cohort generalizability of deep and conventional machine learning for MRI-based diagnosis and prediction of Alzheimer’s disease

This work validates the generalizability of MRI-based classification of Alzheimer’s disease (AD) patients and controls (CN) to an external data set and to the task of prediction of conversion to AD in individuals with mild cognitive impairment (MCI).We used a conventional support vector machine (SVM) and a deep convolutional neural network (CNN) approach based on structural MRI scans that underwent either minimal pre-processing or more extensive pre-processing into modulated gray matter (GM) maps. Classifiers were optimized and evaluated using cross-validation in the Alzheimer’s Disease Neuroimaging Initiative (ADNI; 334 AD, 520 CN). Trained classifiers were subsequently applied to predict conversion to AD in ADNI MCI patients (231 converters, 628 non-converters) and in the independent Health-RI Parelsnoer Neurodegenerative Diseases Biobank data set. From this multi-center study representing a tertiary memory clinic population, we included 199 AD patients, 139 participants with subjective cognitive decline, 48 MCI patients converting to dementia, and 91 MCI patients who did not convert to dementia.AD-CN classification based on modulated GM maps resulted in a similar area-under-the-curve (AUC) for SVM (0.940; 95%CI: 0.924–0.955) and CNN (0.933; 95%CI: 0.918–0.948). Application to conversion prediction in MCI yielded significantly higher performance for SVM (AUC = 0.756; 95%CI: 0.720-0.788) than for CNN (AUC = 0.742; 95%CI: 0.709-0.776) (p<0.01 for McNemar’s test). In external validation, performance was slightly decreased. For AD-CN, it again gave similar AUCs for SVM (0.896; 95%CI: 0.855–0.932) and CNN (0.876; 95%CI: 0.836–0.913). For prediction in MCI, performances decreased for both SVM (AUC = 0.665; 95%CI: 0.576-0.760) and CNN (AUC = 0.702; 95%CI: 0.624-0.786). Both with SVM and CNN, classification based on modulated GM maps significantly outperformed classification based on minimally processed images (p=0.01).Deep and conventional classifiers performed equally well for AD classification and their performance decreased only slightly when applied to the external cohort. We expect that this work on external validation contributes towards translation of machine learning to clinical practice.

Deep Feature Selection and Decision Level Fusion for Lungs Nodule Classification

The existence of pulmonary nodules exhibits the presence of lung cancer. The Computer-Aided Diagnostic (CAD) and classification of such nodules in CT images lead to improve the lung cancer screening. The classic CAD systems utilize nodule detector and feature-based classifier. In this work, we proposed a decision level fusion technique to improve the performance of the CAD system for lung nodule classification. First, we evaluated the performance of Support Vector Machine (SVM) and AdaBoostM2 algorithms based on the deep features from the state-of-the-art transferable architectures (such as; VGG-16, VGG-19, GoogLeNet, Inception-V3, ResNet-18, ResNet-50, ResNet-101 and InceptionResNet-V2). Then, we analyzed the performance of SVM and AdaBoostM2 classifier as a function of deep features. We also extracted the deep features by identifying the optimal layers which improved the performance of the classifiers. The classification accuracy is increased from 76.88% to 86.28% for ResNet-101 and 67.37% to 83.40% for GoogLeNet. Similarly, the error rate is also reduced significantly. Moreover, the results showed that SVM is more robust and efficient for deep features as compared to AdaBoostM2. The results are based on 4-fold cross-validation and are presented for publicly available LUNGx challenge dataset. We showed that the proposed technique outperforms as compared to state-of-the-art techniques and achieved accuracy score was 90.46 ± 0.25%.

Application on Damage Types Recognition of Civil Aeroengine Based on SVM Optimized by DMPSO

In order to recognize the damage types of aeroengine automatically and Reliably, enhance the support capability of aeroengine maintenance, the feature extraction method based on color moments and gray level co-occurrence matrix (GLCM) is proposed to construct the feature database of the aeroengine damage images. The support vector machine (SVM) is utilized as intelligent classifier for damages recognition. Meanwhile, a double-mutations particles swarm optimization (DMPSO) algorithm is designed to optimize the kernel parameter and penalty factor for guaranteeing the recognition performance of SVM. Finally, the feature databases are constructed by different feature methods according to actual four damage types of a certain aeroengine, and then the proposed SVM optimized by DMPSO is used to compare with back propagation (BP) network, extreme learning machine (ELM) network, and k-nearest neighborhood (k-NN). The recognition results have proven the proposed feature extraction method is more suitable for aeroengine damage recognition. Meanwhile, the comparison results have demonstrated the optimized SVM always has better and stable recognition output.

A Novel Gated Recurrent Unit Network Based on SVM and Moth-Flame Optimization Algorithm for Behavior Decision-Making of Autonomous Vehicles

The behavior decision-making algorithm plays an important role in ensuring the safe driving of autonomous vehicles. However, existing behavior decision-making methods lack the capability to cope with future motion uncertainty in traffic, because the historical state of vehicles are not considered. This article proposes a novel driving behavior decision-making method EnMFO-ImGRU based on Gated Recurrent Unit (GRU) and Moth-Flame Optimization algorithm (MFO). Four improvements are proposed in EnMFO-ImGRU. First, to consider the driving information of the vehicles on the road, ImGRU is designed based on a double-layer GRU. Second, to promote decisions accuracy, Support Vector Machine (SVM), which has good performance in classification problems, replaces the softmax classifier to train the output of the ImGRU. Third, to promote the classification capability of SVM, MFO is introduced to optimize the key parameters that affect the performance of SVM. Finally, to promote the optimization capability of MFO, we propose the Enhanced Moth-Flame Optimization algorithm (EnMFO). A new position updating method is proposed in EnMFO. The experimental results on the NGSIM dataset show that EnMFO-ImGRU brings higher accuracy than existing methods for the behavior decision-making results of autonomous vehicles.

Multiple power quality event detection and classification using a modified S-transform and WOA tuned SVM classifier

In this paper, a novel method for classification of power quality events is illustrated. 15 types of power quality events consisting of single and multi-stage disturbances are considered for study. A database of the synthetic PQ events is generated in MATLAB using mathematical models. The generated signals are passed through a novel Modified Stockwell transform consisting of second order gaussian window which provides the ST matrix. From the ST matrix various statistical features such as energy and standard deviation of the magnitude and phase contour are extracted and given as input to Support Vector Machine (SVM). Furthermore, to improve the performance of SVM, a novel meta-heuristic technique called Whale Optimization Algorithm (WOA) is used to tune the hyper parameters of the SVM classifier. The performance of the proposed method is analyzed under noisy and noiseless conditions. It is observed that WOA tuned SVM provides improved classification accuracy than other widely used meta-heuristic optimization algorithms such as Particle Swarm Optimization (PSO) tuned SVM and Genetic Algorithm (GA) tuned SVM. Further, two novel circuits for generation of sag, swell and interrupt are developed and the proposed technique is validated on real time signals obtained from the circuits.

Intrusion Detection Systems, Issues, Challenges, and Needs

Intrusion detection systems (IDSs) are one of the promising tools for protecting data and networks; many classification algorithms, such as neural network (NN), Naive Bayes (NB), decision tree (DT), and support vector machine (SVM) have been used for IDS in the last decades. However, these classifiers is not working well if they applied alone without any other algorithms that can tune the parameters of these classifiers or choose the best sub set features of the problem. Such parameters are C in SVM and gamma which effect the performance of SVM if not tuned well. Optimization algorithms such as genetic algorithm (GA), particle swarm optimization (PSO) algorithm , ant colony algorithm, and many other algorithms are used along with classifiers to improve the work of these classifiers in detecting intrusion and to increase the performance of these classifiers. However, these algorithms suffer from many lacks especially when apply to detect new type of attacks, and need for new algorithms such as JAYA algorithm, teaching learning-based optimization algorithm (TLBO) algorithm is arise. In this paper, we review the classifiers and optimization algorithms used in IDS, state their strength and weaknesses, and provide the researchers with alternative algorithms that could be use in the field of IDS in future works.

A Multidimensional Feature Extraction and Selection Method for ECG Arrhythmias Classification

This article presents a novel multidimensional feature extraction and selection method for electrocardiogram (ECG) arrhythmias classification. First, a novel multi-interval symmetrized dot pattern method (MSDP) is proposed to extract image shape features, which are combined with time-frequency features, morphological features and RR interval features to form a new feature set. Then a novel average weight based ReliefF method (WReliefF) is proposed to integrate with genetic algorithm (GA) and support vector machine (SVM) and form a feature selection method named as WReliefF-GA-SVM. The WReliefF-GA-SVM is a hybrid type feature selection method, and it can be divided into Filter stage and Wrapper stage to select features. In Filter stage, the algorithm uses WReliefF to sort the features according to the average weight. In Wrapper stage, GA and SVM are combined, and the classification performance of SVM is used as the fitness function to search the best feature subsets. Finally, the five types of ECG beats from the MIT-BIH arrhythmia database are classified by a one against all (OAA) SVM model. The results show that the proposed method achieves average sensitivity, specificity and accuracy of 99.42%, 99.83%, and 99.74%, respectively.