Vector Machine Classification Algorithm Research Articles

Design and Exploration of College English Reading, Writing, and Translation Teaching Classroom Based on Machine Learning

The aims are to ameliorate the dull classroom atmosphere and unsatisfactory teacher-student interaction and cultivate students’ reading, writing, and translation (R-W-T) English proficiency level (EPL). Specifically, this paper designs an optimized polynomial kernel-based support vector machine (SVM) classification algorithm based on the machine learning (ML) theory. Firstly, this paper expounds on the correlation between ML and teaching classrooms to analyze the optimization direction in the application of ML in classroom teaching. Afterward, SVM and RF algorithms are selected for data normalization optimization, and their optimal hyperparameters are analyzed. Consequently, an experiment is designed to compare the classification accuracy of the algorithms before and after optimization, and the polynomial kernel-based SVM algorithm is proved to present the most remarkable improvement and accuracy after optimization, which is as high as 95.23% or 66.96% improvement. Therefore, the polynomial kernel-based SVM algorithm is chosen for the college English (R-W-T) classroom-oriented human pose recognition (HPR) system. Thus, English teachers can better grasp the students’ psychological state and classroom atmosphere and ameliorate unsatisfactory teacher-student interaction in students’ English (R-W-T). The proposal plays a positive role in cultivating college students’ strong (R-W-T) EPL, which is of great significance in improving the English classroom.

Feature Expansion Word2Vec for Sentiment Analysis of Public Policy in Twitter

Social media users, especially on Twitter, can freely express opinions or other information in the form of tweets about anything, including responding to a public policy. In a written tweet, there is a limit of 280 characters per tweet and this allows for problems such as vocabulary mismatches. Therefore, in this study, the feature expansion Word2vec method was applied to overcome when the vocabulary mismatches occur. This study develops and compares the Twitter sentiment analysis system using the feature expansion Word2vec method with the Logistic Regression (LR) and Support Vector Machine (SVM) classification algorithms and the system without the feature expansion Word2Vec method. The results of this study, the feature expansion Word2Vec method on the SVM classification algorithm succeeded in increasing the system accuracy up to 0,99% with an accuracy value of 78,99%.

Arteriovenous Fistula Flow Dysfunction Surveillance: Early Detection Using Pulse Radar Sensor and Machine Learning Classification

Vascular Access (VA) is often referred to as the “Achilles heel” for a Hemodialysis (HD)-dependent patient. Both the patent and sufficient VA provide adequacy for performing dialysis and reducing dialysis-related complications, while on the contrary, insufficient VA is the main reason for recurrent hospitalizations, high morbidity, and high mortality in HD patients. A non-invasive Vascular Wall Motion (VWM) monitoring system, made up of a pulse radar sensor and Support Vector Machine (SVM) classification algorithm, has been developed to detect access flow dysfunction in Arteriovenous Fistula (AVF). The harmonic ratios derived from the Fast Fourier Transform (FFT) spectrum-based signal processing technique were employed as the input features for the SVM classifier. The result of a pilot clinical trial showed that a more accurate prediction of AVF flow dysfunction could be achieved by the VWM monitor as compared with the Ultrasound Dilution (UD) flow monitor. Receiver Operating Characteristic (ROC) curve analysis showed that the SVM classification algorithm achieved a detection specificity of 100% at detection thresholds in the range from 500 to 750 mL/min and a maximum sensitivity of 95.2% at a detection threshold of 750 mL/min.

An Enhanced Twitter Corpus for the Classification of Arabic Speech Acts

Twitter has gained wide attention as a major social media platform where many topics are discussed on daily basis through millions of tweets. A tweet can be viewed as a speech act (SA), which is an utterance for presenting information, hiding indirect meaning, or carrying out an action. According to SA theory, SA can represent an assertion, a question, a recommendation, or many other things. In this paper, we tackle the problem of constructing a reference corpus of Arabic tweets for the classification of Arabic speech acts. We refer to this corpus as the Arabic Tweets Speech Act Corpus (ArTSAC). It is an enhancement of a modern standard Arabic (MSA) tweet corpus of speech acts called ArSAS. ArTSAC is more advantageous than ArSAS in terms of its richness of annotated features. The goal of ArTSAC is twofold: Firstly, to understand the purpose and intention of tweets which act in accordance with the SA theory, and hence positively influencing the development of many natural language processing (NLP) applications. Secondly, as a future goal, to be used as a benchmark annotated dataset for testing and evaluating state-of-the-art Arabic SA classification algorithms and applications. ArTSAC has been put in practice to classify Arabic tweets containing speech acts using the Support Vector Machine (SVM) classification algorithm. The results of the experiments show that the enhanced ArTSAC corpus achieved an average precision of 90.6% and an F-score of 89.6%. Substantially it outperformed the results of its predecessor ArTSAC corpus.

Intelligent Symbiotic Relay Selection Technique for 5G Networks

The growing demand for bandwidth and spectrum has inspired the ongoing efforts to establish the future 5G network supporting vertical sectors such as cyber-physical systems (CPS). Cooperative communication is one of the requisite techniques to improve coverage, network capacity and reduce power consumption in the network. In this paper, a symbiotic two-phase intelligent transmission is considered. The first phase occurs between the source and the candidate relays, and involves the selection of a set of “reliable relays”. The second phase occurs between the reliable relays and the destination, and involves the selection of the “best relay” for transmission. Dynamic relay selection using k-means clustering is used to detect the most significant correlation between all the channel state information (CSI) attributes in the system. The work identified the reliable relays while reducing the number of relay nodes for the second transmission phase. Contextual scenarios are created with typical network configuration using three geographical locations Coventry, Birmingham and London. An experimental validation is done with Omnet++ environment for the scenarios of three geographical locations. A natural grouping of mobile users is carried out leveraging the relay capabilities. The results are validated using support vector machine (SVM) classification algorithm. Considering urban environment deployment of relay nodes, metrics such as signal-to-noise-plus-interference ratio (SINR), attenuation, signal to noise ratio (SNR), link quality, k-means clustering, accuracy, and root mean square error (RMSE) are investigated for the Direct-2-Direct (D2D) capable relays. It was observed that the proposed technique both outperforms the other fixed-parameter relay selection techniques and improves with larger datasets unlike the other techniques.

Annotation of medical images using integral images and Gabor coefficients

Image annotation is one of the aspects of automatic classification algorithms. In this paper, we propose a learning-based algorithm for automatic medical image annotation based on local appearance cues by using integral images, Gabor transforms, AdaBoost and support vector machine (SVM) classification algorithm. The algorithm starts with a number of landmarks. By landmarks, we generate special sub-patches and extract features using Gabor transforms. In fact, Gabor filter is applied to integral image. AdaBoost in most cases can make a decision about the instances at this stage. To solve the misclassified cases, we combine similar classes and employ SVM/AdaBoost to recognise close instances. This approach is evaluated on large-scale radiograph images and several experiments have been done. At first, we specify five classes with the most number of images in ImageCLEF2008. Then, the proposed method is used for position identification task. At the end, our approach achieved the acceptable accuracies for multiclass radiograph annotation task, when compared to other algorithms. We obtain the accuracy of 95% for multiclass radiograph annotation and >99% for position identification task.

Identification of unknown operating system type of Internet of Things terminal device based on RIPPER

Due to the vast popularity of sensors, cloud computing, mobile computing, and intelligent devices, the Internet of Things has seen tremendous growth in recent years. Operating system type recognition is the core technology of network security assessment. Due to inherit security problems of Internet of Things such as the situation of risk and threat of information, the operating system recognition seeks research attention for Internet of Things network security. In view of the current identification method of active operating system, it is prone to be detected by intrusion detection system. The operating system identification technology based on transmission control protocol/Internet protocol fingerprint library is more complicated than to distinguish the operating system types of unknown fingerprints. In this work, a passive operating system identification method based on RIPPER model is proposed. Also, it is compared with the existing support vector machine and C45 decision tree classification algorithms. Experiments reveal that RIPPER-based algorithm has better recognition accuracy and recognition efficiency.

Dimension Reduction and Defect Recognition of Strip Surface Defects Based on Intelligent Information Processing

Feature extraction and classification recognition are very important for surface defect recognition of strip steel. Traditional methods of feature extraction and defect recognition cannot take into account both the recognition accuracy and the running time. In order to solve the above two problems, a novel feature extraction and recognition method are proposed in this paper. On the one hand, a semi-supervised principal component analysis and locality preserving projection manifold learning algorithm are proposed to reduce features dimension. On the other hand, the particle swarm optimization–second-order cone programming–multi-kernel relevance vector machine (PSO–SOCP–MKRVM) classification algorithm is proposed in this paper, which is used to improve efficiency and accuracy. The experimental results show that the novel recognition method can not only improve the recognition accuracy, but also meet the real-time requirements of online detection.

Fatigue crack monitoring of aerospace structure based on lamb waves and binary tree support vector machines

To overcome the difficulty in identifying the fatigue crack in key parts of aerospace structure, a kind of methods aimed to monitor the crack length based on matching pursuit (MP) method and binary tree support vector machines (BT-SVM) classification algorithm was developed. In this method, Lamb wave signals were decomposed into a linear combination of several Chirplet atoms by MP method, and then the matching parameters were extracted as feature vectors for training and testing in BT-SVM classification algorithm. At the same time, a lug joint model was created with a certain ratio and the effect of crack extension on Lamb wave signals propagation was studied. At last, fatigue loading experiments were carried out in lug joints and tail reinforced frames of aircraft. The results showed that this method can monitor the length of fatigue crack effectively, which presents a new approach for monitoring the fatigue crack.

The facial expression of schizophrenic patients applied with infrared thermal facial image sequence

BackgroundSchizophrenia is a neurological disease characterized by alterations to patients’ cognitive functions and emotional expressions. Relevant studies often use magnetic resonance imaging (MRI) of the brain to explore structural differences and responsiveness within brain regions. However, as this technique is expensive and commonly induces claustrophobia, it is frequently refused by patients. Thus, this study used non-contact infrared thermal facial images (ITFIs) to analyze facial temperature changes evoked by different emotions in moderately and markedly ill schizophrenia patients.MethodsSchizophrenia is an emotion-related disorder, and images eliciting different types of emotions were selected from the international affective picture system (IAPS) and presented to subjects during ITFI collection. ITFIs were aligned using affine registration, and the changes induced by small irregular head movements were corrected. The average temperatures from the forehead, nose, mouth, left cheek, and right cheek were calculated, and continuous temperature changes were used as features. After performing dimensionality reduction and noise removal using the component analysis method, multivariate analysis of variance and the Support Vector Machine (SVM) classification algorithm were used to identify moderately and markedly ill schizophrenia patients.ResultsAnalysis of five facial areas indicated significant temperature changes in the forehead and nose upon exposure to various emotional stimuli and in the right cheek upon evocation of high valence low arousal (HVLA) stimuli. The most significant P-value (lower than 0.001) was obtained in the forehead area upon evocation of disgust. Finally, when the features of forehead temperature changes in response to low valence high arousal (LVHA) were reduced to 9 using dimensionality reduction and noise removal, the identification rate was as high as 94.3%.ConclusionsOur results show that features obtained in the forehead, nose, and right cheek significantly differed between moderately and markedly ill schizophrenia patients. We then chose the features that most effectively distinguish between moderately and markedly ill schizophrenia patients using the SVM. These results demonstrate that the ITFI analysis protocol proposed in this study can effectively provide reference information regarding the phase of the disease in patients with schizophrenia.

A novel method for urban area extraction from VIIRS DNB and MODIS NDVI data: a case study of Chinese cities

ABSTRACTMapping urban areas at the regional and global scales has been used in ecology, environment, sociology, and other subjects. Recently, it has become increasingly popular to extract urban areas from night-time light remote-sensing data. In this article, we reported an alternative method to extract information of urban areas from VIIRS Day/Night Band (DNB) and MODIS normalized differential vegetation index (NDVI) data based on the adaptive mutation particle swarm optimization (AMPSO) algorithm and the Support Vector Machine (SVM) classification algorithm. This method was validated using the urban areas of nine Chinese cities classified from Landsat Enhanced Thematic Mapper (ETM+) images by object-oriented image classification technology. We demonstrated that this new method for urban area extraction had a good classification coherency with the Landsat8 OLI result. In addition, it is more robust than other classification methods, and can be used to characterize the inter-urban texture as well.

Analysis of Various Types of Bugs in the Object Oriented Java Script Language Coding

Objective: The main goal of this work is to categorize and understand the various types of bugs present in the Object oriented java scripting language. By finding these faults one can find out the nature of faults which causes the run time failure of programs. Methods: There are various types of bugs present in the OOJS environment which needs to be categorized well to understand the nature of fault which has occurred in the OOJS language. By doing so, web pages can be prevented from the functioning failure and can cause the generation of more user flexible environment. Fault Localization is the approach to analyse and detect the place of faults present in the java scripting language. Transductive Support Vector Machine (TSVM) classification algorithm is introduced to categorize the faults into various sub types by classifying them based on types of bugs. This classification is done on the bug report data set which was created by using the fault localization approach. Findings: The experimental test conducted proves that the proposed approach named TSVM based categorization of faults can detect the faults efficiently which can be used further for error detection. The proposed approach improves in its performance in terms of improved accuracy detection by categorizing the faults correctly through which faults present in the programming language can be done efficiently. Conclusion: From the findings it can be concluded that the proposed approach improved in its performance in terms of improved accuracy.

A novel multi-parameter support vector machine for image classification

The support vector machine (SVM) classification algorithm has received increasing attention in recent years in remote sensing for land-cover classification. However, it is well known that the performance of the SVM is sensitive to the choice of parameter settings. The traditional single optimized parameter SVM (SOP-SVM) attempts to identify globally optimized parameters for multi-class land-cover classification. In this article, a novel multi-parameter SVM (MP-SVM) algorithm is proposed for image classification. It divides the training set into several subsets, which are subsequently combined. Based on these combinations, sub-classifiers are constructed using their own optimum parameters, providing votes for each pixel with which to construct the final output. The SOP-SVM and MP-SVM were tested on three pilot study sites with very high, high, and low levels of landscape complexity within the Sanjiang Plain – a typical inland wetland and freshwater ecosystem in northeast China. A high overall accuracy of 82.19% with kappa coefficient (κ) of 0.80 was achieved by the MP-SVM in the very high-complexity landscape, statistically significantly different (z-value = 3.77) from the overall accuracy of 72.50% and κ of 0.69 produced by the traditional SOP-SVM. Besides, for the moderate-complexity landscape a significant increase in accuracy was achieved (z-value = 2.44), with overall accuracy of 84.03% and κ of 0.80 compared with an overall accuracy 76.05% and κ of 0.71 for the SOP-SVM. However, for the low-complexity landscape the MP-SVM was not significantly different from the SOP-SVM (z-value = 0.80). Thus, the results suggest that the MP-SVM method is promising for application to very high and high levels of landscape complexity, differentiating complex land-cover classes that are spectrally mixed, such as marsh, bare land, and meadow.

Identification of mobility changes induced by deoxyribo nucleic acid in the inversion regions of ultralong single walled carbon nanotube field-effect transistor using support vector machine and multi-layer perceptron

Using ethanol based chemical vapour deposition (CVD), ultralong single walled carbon nanotubes (SWCNTs) were consistently grown in the centimetre scale, and a number of long-channel carbon nanotube field-effect transistors (CNTFETs) were fabricated. Here, we extracted the field-effect mobility of millimetre-long CNTFETs from the strong inversion region and near-threshold region respectively before and after the addition of deoxyribo nucleic acid (DNA). Support vector machine (SVM) and multi-layer perceptron (MLP) classification algorithms were used to discriminate between the bare device mobility and the DNA induced mobility data in the two regions. SVM classification yielded an average accuracy, sensitivity, and specificity of 97.81%, 97.66%, and 95.38% respectively. On the other hand, MLP based classification resulted in an average accuracy, sensitivity, and specificity of 92.10%, 95.07%, and 90.70%, respectively.

Vehicle Classification Algorithm Based on Fuzzy SVM Models

As a typical binary classifier, its an inseparable sample problem about the Support Vector Machine (SVM) when processing the classification of the multi-class vehicle models. Since the SVM can not estimate the effect size of the samples classification accurately, and then reduces the classification generalization ability. In this paper, a fuzzy Support Vector Machine (FSVM) classification algorithm is applied to vehicle classification. According to the difference of the contribution which the vehicle characteristics make to the classification, the appropriate degree of membership is given, and the algorithm improves the vehicle models classification ability of the traditional SVM effectively. The experimental results show that the new method, compared with the existing vehicle classification method, is feasible, effective, and with a high classification accuracy

Recognition of Different Modes of Human Thinking during Processing Visual Images of Histograms and Scenes

To investigate the different modes of human thinking, we designed an eye tracking experiment during people recognized two category images of histograms and scenes, and used the support vector machine (SVM) classification algorithm to classify these eye movement data. The results of statistical analysis showed that there were significant differences in saccade distance and pupil diameter between these two category images. By the feature selection, normalization of data preprocessing, and SVM classification, the results of classification analysis showed that there was a better performance on the classification of the histograms and scenes. These results suggest we can identify the modes of human thinking through the SVM classification methods based on the eye movement data.

An 18-gene blood-based IVD test for colorectal cancer early detection with high sensitivity and specificity.

3551 Background: Colorectal cancer (CRC) is often curable particularly diagnosed at early stages. Different screening strategies are in place in various counties. However, the compliance rate with current CRC screening recommendations remains poor. A simple blood-based IVD test would represent an interesting alternative. We previously reported a 100-gene signature for early detection of CRC using microarray technology. Based on this result, we have developed a new blood test with real-time PCR technology. Methods: 144 CRC cases and 153 Controls were enrolled. A total of 52 genes were selected as candidate markers to be transferred from microarray to real-time PCR. The gene expression profiles from 100 CRC cases and 100 Controls were used as a training set for signature discovery. Signature identification process was conducted by Minimum Redundancy Maximum Relevance (mRMR) feature selection method and Support Vector Machine (SVM) classification algorithm under the Leave-One-Out Cross Validation (LOOCV) framework. The optimal size of signature was determined by a stepwise inclusion selection procedure which started with one gene and repeatedly added genes one by one until all genes were included. The performance of each N-gene signature by the LOOCV was estimated and used to determine the best size of signature. Results: The gene expression profiles measured by microarray and real-time PCR technology were highly comparable and resulted in a significant correlation in term of fold change ratio between CRC cases and Controls (Pearson’s Correlation Coefficient = 0.95). An 18-gene signature was identified and validated in an independent validation set with 44 CRC cases and 53 Controls. The performance of 18-gene signature in the validation set reached 88.7% accuracy (95%CI: 0.80-0.94), 88.6% sensitivity (95%CI: 0.75-0.96), and 88.7% specificity (95%CI: 0.76-0.95). Conclusions: Our new results demonstrated the feasibility of technology transfer from microarray to real-time PCR. The blood test could be offered to individuals who are unwilling or unable to undergo colonoscopy to increase the CRC screening compliance rate. Meanwhile, its high specificity could also help to avoid unnecessary colonoscopies.

Looking Beyond Historical Patient Outcomes to Improve Clinical Models

Conventional algorithms for modeling clinical events focus on characterizing the differences between patients with varying outcomes in historical data sets used for the model derivation. For many clinical conditions with low prevalence and where small data sets are available, this approach to developing models is challenging due to the limited number of positive (that is, event) examples available for model training. Here, we investigate how the approach of developing clinical models might be improved across three distinct patient populations (patients with acute coronary syndrome enrolled in the DISPERSE2-TIMI33 and MERLIN-TIMI36 trials, patients undergoing inpatient surgery in the National Surgical Quality Improvement Program registry, and patients undergoing percutaneous coronary intervention in the Blue Cross Blue Shield of Michigan Cardiovascular Consortium registry). For each of these cases, we supplement an incomplete characterization of patient outcomes in the derivation data set (uncensored view of the data) with an additional characterization of the extent to which patients differ from the statistical support of their clinical characteristics (censored view of the data). Our approach exploits the same training data within the derivation cohort in multiple ways to improve the accuracy of prediction. We position this approach within the context of traditional supervised (2-class) and unsupervised (1-class) learning methods and present a 1.5-class approach for clinical decision-making. We describe a 1.5-class support vector machine (SVM) classification algorithm that implements this approach, and report on its performance relative to logistic regression and 2-class SVM classification with cost-sensitive weighting and oversampling. The 1.5-class SVM algorithm improved prediction accuracy relative to other approaches and may have value in predicting clinical events both at the bedside and for risk-adjusted quality of care assessment.

The learning performance of support vector machine classification based on Markov sampling

The previously known frameworks describing the consistency of support vector machine classification (SVMC) algorithm are usually based on the assumption of independent and identically distributed (i.i.d.) samples. In this paper we go far beyond these classical frameworks by studying the consistency of SVMC algorithm with uniformly ergodic Markov chain samples based on linear prediction models. We establish the bound on the consistency of SVMC algorithm with uniformly ergodic Markov chain samples, and show that SVMC algorithm with uniformly ergodic Markov chain samples is consistent. Inspired by the idea from Markov chain Monto Carlo (MCMC) methods, we introduce a new Markov sampling algorithm for classification to generate uniformly ergodic Markov chain samples from large data set, and present numerical studies on simulated data and benchmark repository using SVMC algorithm.

KFDA and clustering based multiclass SVM for intrusion detection

To improve the classification accuracy and reduce the training time, an intrusion detection technology is proposed, which combines feature extraction technology and multiclass support vector machine (SVM) classification algorithm. The intrusion detection model setup has two phases. The first phase is to project the original training data into kernel fisher discriminant analysis (KFDA) space. The second phase is to use fuzzy clustering technology to cluster the projected data and construct the decision tree, based on the clustering results. The overall detection model is set up based on the decision tree. Results of the experiment using knowledge discovery and data mining (KDD) from 99 datasets demonstrate that the proposed technology can be an an effective way for intrusion detection.