Multi-classification Model Research Articles

ECG-Signal Multi-Classification Model Based on Squeeze-and-Excitation Residual Neural Networks

Accurate electrocardiogram (ECG) interpretation is crucial in the clinical ECG workflow because it is most likely associated with a disease that can cause major problems in the body. In this study, we proposed an ECG-signal multi-classification model using deep learning. We used a squeeze-and-excitation residual network (SE-ResNet), which is a residual network(ResNet) with a squeeze-and-excitation block. Experiments were performed for seven different types of lead-II ECG data obtained from the Korea University Anam Hospital in South Korea. These seven types are normal sinus rhythm, atrial fibrillation, atrial flutter, sinus bradycardia, sinus tachycardia, premature ventricular contraction and first-degree atrioventricular block. We compared the SE-ResNet with a ResNet, as a baseline model, for various depths of layer (18/34/50/101/152). We confirmed that the SE-ResNet had better classification performance than the ResNet, for all layers. The SE-ResNet classifier with 152 layers achieved F1 scores of 97.05% for seven-class classifications. Our model surpassed the baseline model, ResNet, by +1.40% for the seven-class classifications. For ECG-signal multi-classification, considering the F1 scores, the SE-ResNet might be better than the ResNet baseline model.

A Novel Method for Validating Multi-Classifiers. A Case Study for ICF-Based Health Status Classification.

In this paper, we propose a novel method for the validation of a multi-classification model according to the intended use and aim of a device for health status classification and the clinical needs of the practitioners involved.

Dynamic and Probabilistic Multi-class Prediction of Tunnel Squeezing Intensity

Tunnel squeezing is a time-dependent process that typically occurs in weak or over-stressed rock masses, significantly influencing the budget and time of tunnel construction. This paper presents a new framework to probabilistically predict the potential squeezing intensity and to dynamically update the prediction during construction based on the sequentially revealed ground information. An extensively well-documented database, which contains quantitative data from 154 squeezing sections with 95 unpublished inventories is established. A Decision Tree method is employed to train a probabilistic multi-classification model to predict the tunnel squeezing intensity. The trained classifier is then integrated with a Markovian geologic model, which features embedded Bayesian updating procedures, to achieve a dynamic prediction on the state probabilities of the geologic parameter within the model and the resulting squeezing intensity during excavation. An under-construction tunnel case—Miyaluo #3 tunnel—is used to illustrate the proposed framework. Results show that the Decision Tree classifier, as opposed to other black-box models, is easy to be interpreted. It provides reliable predictive accuracy while leading to insights into the understanding of the squeezing problem. The strength-stress ratio (SSR) is suggested to be the most important factor. Moreover, the implementation of the updating procedures is efficient since only a simple field test (e.g. Point Load index or Schmidt rebound index) is required. Multiple rounds of predictions within the updating process allow different levels of prediction, for example long-range, short-term, or immediate, to be extracted as useful information towards the decision-making of construction operations. Therefore, this framework can serve as a pragmatic tool to assist the selection of optimal primary-support and other construction strategies based on the potential squeezing risk.

A Multi-Fault Detection Method With Improved Triplet Loss Based on Hard Sample Mining

Fault detection plays an essential role in the power generation of wind turbines (WTs). However, most of the present fault detection methods are designed to detect a certain type of fault. When these methods are used to detect multiple fault types, their accuracies tend to decline. In this paper, an effective multi-fault detection method based on improved triplet loss is proposed. First, a hard sample mining method is presented based on WT data for the first time, so that the samples which are more suitable for triplet training are selected. Second, triplet training is improved by applying a new mapping layer and optimizing the loss function. Third, based on the improved triplet training, a deep learning model is trained to map the original samples into feature vectors in an embedding space, so that the distance between two samples in the same class is shortened whereas that of different classes is lengthened. Then, a multi-classification model is trained to distinguish the feature vectors, so that the faults of different types can be detected. Finally, four groups of experiments are conducted using real WT SCADA data collected from a wind farm in northern China. The experiment results show that the proposed method is effective.

A Bearing Fault Diagnosis Method Based on Dictionary Learning and Parameter-Optimized Support Vector Machine

Aiming at the problem of fault diagnosis, a novel method based on dictionary learning and parameter-optimized Support Vector Machine (SVM) was proposed in this paper and applied it to bearing fault diagnosis. Firstly, the collected bearing fault signals are transformed into gray images after data processing. Then, using dictionary learning, the gray images are denoised and output them as signal data. Finally, the SVM multi-classification model obtained by using Grid Search (GS) algorithm to optimize penalty parameter c and kernel function parameter g is used to classify and identify the fault type. This paper is based on data from Case Western Reserve University Bearing Center for experimental verification. The results show that the proposed model can continuously achieve the accuracy of 100% in the process of bearing fault diagnosis in different environments, which proves that the proposed method can accurately and effectively realize the fault diagnosis classification of bearings.

Using Vehicle Interior Noise Classification for Monitoring Urban Rail Transit Infrastructure.

This study developed a multi-classification model for vehicle interior noise from the subway system, collected on smartphones. The proposed model has the potential to be used to analyze the causes of abnormal noise using statistical methods and evaluate the effect of rail maintenance work. To this end, first, we developed a multi-source data (audio, acceleration, and angle rate) collection framework via smartphone built-in sensors. Then, considering the Shannon entropy, a 1-second window was selected to segment the time-series signals. This study extracted 45 features from the time- and frequency-domains to establish the classifier. Next, we investigated the effects of balancing the training dataset with the Synthetic Minority Oversampling Technique (SMOTE). By comparing and analyzing the classification results of importance-based and mutual information-based feature selection methods, the study employed a feature set consisting of the top 10 features by importance score. Comparisons with other classifiers indicated that the proposed XGBoost-based classifier runs fast while maintaining good accuracy. Finally, case studies were provided to extend the applications of this classifier to the analysis of abnormal vehicle interior noise events and evaluate the effects of rail grinding.

Seismic Waveform Classification Based on Multi-classification Model Weighted Voting Method

The earthquake event is caused by the activity of the interior earth, which has the potential and can create great damage. Thus, in big data era, it is necessary to fast, efficient and universal automatic recognition and classification of seismic waveform. A well-performed recognition and classification algorithm should be sensitive to small and weak events with various waveform shapes, robust to background noise and non-seismic signals, and effective for processing a large number of data in different areas. In this paper, we introduce the weighted voting method in ensemble learning, taking 50000 seismic signals monitored by the capital circle and its nearby stations as training and testing data, we learn the main time-frequency characteristics of seismic positions in the three-component seismic signal data through the sub-model, and then identify and classify them. Finally, the classification score of the sub-model is weighted to confirm the final event type. Compared with single machine learning models, our result shows improved recognition and classification performance and reduced uncertainty of ensembles.

A Multiclassification Model of Sentiment for E-Commerce Reviews

Consumer reviews are important information that reflects the quality of E-commerce goods and services and their existing problems after shopping. Due to the possible differences in consumers' experiences with goods and service quality, consumer reviews can involve multiple-aspect expressions of emotions or opinions. This may result in attitudes expressed by a consumer in the same review sometimes having a variety of emotions. We introduce a sentiment multiclassification method based on a directed weighted model. The model represents the sentiment entity vocabulary as the sentiment nodes and represents the relation between nodes as the directed weighted link. The sentiment entity vocabulary is the entity with attributes, which can express sentiment meaning in related reviews. Directed weighted links represent the sentiment similarity between two nodes of entities with attributes and determined by the direct correlation calculation between them. The paths are all connected directed links from one node to another, which are composed of several nodes and links with close sentiment similarity. Then, we can establish a directed weighted model concerning the sentiments. Directed weighted links having similar sentiment relations with each other may constitute a directed weighted path. There are several directed weighted paths from a start node to the end nodes of the sentiment entity vocabulary in the directed weighted model. Each different path is a different sentiment expression, which represents a different sentiment type. The different sentiment classifications can be obtained through the restriction of path length. Experiments and analysis of the results show that the sentiment multiclassification model based on the directed weighted model proposed in this paper can classify the review sentiments according to different limited threshold rules. Comprehensive analysis indicates the classification results have good accuracy and high efficiency.

Distributed Adaptive Clustering Based on Maximum Correntropy Criterion Over Dynamic Multi-Task Networks

This paper focuses on the problem of distributed adaptive estimation over dynamic multi-task networks, where a set of nodes is required to collectively estimate some parameters of interest from noisy measurements. Besides, since nodes in the network are constrained by communication power consumption and external interference in a non-stationary environment, the objective pursued by the node is prone to change or abnormality. The problem is worth considering in several contexts including multi-target tracking, multi-model classification and heterogeneous network segmentation. We propose a distributed adaptive clustering strategy, which is mainly composed of two procedures: normal task adaptation and the same task cluster. The task anomaly detection based on non-cooperative least-mean-squares (NC-LMS) algorithm and task switching detection based on diffusion maximum correntropy criterion (D-MCC) algorithm are provided. A series of scenarios, such as dynamic network, time-varying tasks and non-stationary (Gaussian and pulse interference) are simulated. We also discuss optimization schemes to design the NC-LMS and D-MCC weights and examine the estimate performance and clustering effects of the proposed algorithm by simulation results.

Reinforcement based Multi-Model Deep Learning Algorithm for Classification

Nowadays researchers are focused on processing the multi-media data for classifying the queries of end users by using search engines. The hybrid combination of a powerful classifier and deep feature extractor are used to develop a robust model, which is performed in a high dimensional space. In this research, a three different types of algorithms are combined to attain a stochastic belief space policy, where these algorithms include generative adversary modelling, maximum entropy Reinforcement Learning (RL) and belief space planning which leads to develop a multi-model classification algorithm. In the simulation framework, different adversarial behaviours are used to minimize the agent's action predictability, which has resulted the proposed method to attain robustness, while comparing with unmodelled adversarial strategies. The proposed reinforcement based Deep Learning (DL) algorithm can be used as multi-model classification purpose. The single neural network algorithm can perform the classification on text data and image data. The RL learns the appropriate belief space policy from the feature extracted information of the text and image data, the belief space policy is generated based on the maximum entropy computation.

Research on Feature Recognition of UAV Acoustic Signal Based on SVM

At present, the analysis of UAV flight acoustic signals is mainly based on traditional speech signal processing methods, and has not been analyzed in depth. According to the flight signal of UAV, combined with the aerodynamic characteristics of UAV, the characteristics of UAV’s acoustic signal are analyzed. The three feature extraction algorithms of pitch period, FFT and Mel Cepstral Coefficient (MFCC) are analyzed and compared. Feature extraction is performed, and a support vector machine (SVM) classification algorithm is applied to perform multi-classification model recognition. The measured and experimental results show that based on SVM classification and recognition, the three feature recognition methods all realize the classification of the model. The comprehensive FFT is the best, and the MFCC is the second. The pitch period is not suitable as the feature extraction method alone.

Fault Diagnosis of High Voltage Circuit Breaker Based on Multi-classification Relevance Vector Machine

The high voltage circuit breaker’s fault as an important form of electrical contact fault in the power system, which is extremely difficult to diagnose under the condition of small fault dataset. This paper proposes a fault diagnosis method based on multi-classification relevance vector machine for high voltage circuit breakers. To make up with the scarcity of the sample fault data in classifying the features of the high voltage circuit breakers, a multi-classification relevance vector machine algorithm is designed on the basis of “One-Against-One” multi-classification model, and tested by public data-sets to verify the good generating performance of this algorithm on small sample data-sets. Then, the time and the currents features are extracted from the closing coil current information of the high voltage circuit breaker to form fault eigenvector. Consequently, numerous two-classification relevance vector machine models were trained and then tested for the optimality of the acquired parameters. The results show that the proposed algorithm can effectively identify many faults of circuit breaker and has better classification accuracy than BP neural network and Support Vector Machine under conditions of small sample data.

Haze Image Classification Method Based on Alexnet Network Transfer Model

In recent years, the serious haze weather frequently appears in China. In order to improve the efficiency of the traffic image dehazing algorithm, it is important to effectively identify whether there is fog and haze in the images. Aiming at the problems of large amount of data sets, low training efficiency, long time and low accuracy in traditional image classification, this paper combines the knowledge of transfer learning and proposes a traffic haze image classification method based on Alexnet network transfer model. First, the image size is normalized and pre-processed. Then, according to the Alexnet network model, the trained multi-classification model is transferred to the haze image binary classification model to form a new Alexnet network transfer model. At last, the network parameters are retrained according to the training parameters in the transfer model, and applied to the haze image classification. Compared with the traditional classification algorithm, the proposed algorithm greatly reduces the amount of data needed for training, and the classification accuracy of the transfer network can reach 97%, which is 9% higher than the traditional SVM algorithm. It can effectively identify haze images.

VSimilar: A high-adaptive VM scheduler based on the CPU pool mechanism

Abstract In a virtualized system, the virtual machine (VM) scheduler plays a key role to the performance promotion of the virtual machine monitor (VMM), a.k.a. hypervisor. The scheduler is responsible for assigning adequate system resources to each VM according to the demands of the VM tenants, which is quite challenging as VM tenants’ demands are quite dynamic and unpredictable. To this end, CPU pool mechanism has been widely adopted as an adaptive solution. However, the CPU pool mechanism still has defficiency in terms of VM classification model and time-slice allocation strategy, as the two strategies have to be effectively utilized for realizing a high-adaptive VM scheduler. In this paper, we thus explore opportunities to improve the CPU pool mechanism and develop a new VM scheduling solution, called vSimilar, which uses VM multi-classification model to more effectively adaptive to the VMs of running different types of tasks at different time. Moreover, by a dynamic time-slice function, vSimilar manages to provide a more efficient resource allocation. The experimental evaluation shows that vSimilar can significantly improve the performance of a VMM, such as Xen. The improvements include 1) a VM server hosted by Xen with vSimilar can reduce nearly 95% of a client’s Ping round-trip time (Ping RTT), 2) vSimilar can help increase about 40% the TCP throughput, and about 20% the UDP throughput, between a Xen-hosted VM server and a client, and 3) vSimilar also increases the page operation rate by nearly 50% for a Xen-hosted VM Web server.

Multi-View Learning With Robust Generalized Eigenvalue Proximal SVM

Multi-view learning mechanism, which enhances learning performance by training multi-model data sets, is a popular filed in recent years. Multi-view generalized eigenvalue proximal support vector machine (MvGSVM), as a most recently proposed classifier, has been shown to be successful in multi-model classification, which incorporates multi-view learning into classical GEPSVM. However, this method is still based on squared L2-norm distance measure, thus its robustness is not guaranteed in the presence of outliers. To address this problem, we propose a robust multi-view GEPSVM based on Lp-norm minimization and Ls-norm maximization. But, the introduction of the Lp-norm and Ls-norm makes the problem different from the generalized eigenvalue problem. So an efficient iterative algorithm is designed to solve this problem, and we also give the proof of convergence of the algorithm. The performances in extensive experiments demonstrate the effectiveness and robustness of the algorithm.

Fault Diagnosis of Variable Load Bearing Based on Quantum Chaotic Fruit Fly VMD and Variational RVM

Under normal circumstances, bearings generally run under variable loading conditions. Under such conditions, the vibration signals of the bearing malfunctions are often nonstationary signals, which are difficult to process effectively. In order to accurately and effectively diagnose the failure types and damage degree of bearings under variable load conditions, an intelligent diagnostic model based on the variational mode decomposition (VMD) of quantum chaotic fruit fly optimization algorithm (QCFOA) and a multiclassification variational relevance vector machine (VRVM) is proposed. First, the key parameters of the VMD are selected using the QCFOA. Secondly, the known bearing fault signal is decomposed by the optimized VMD, and the center frequency and marginal spectral entropy (MSE) of each natural modal component are extracted to construct two‐dimensional MSE. Then, the probit model is used to replace the logistic model, and a simpler and more practical multiclassification model is constructed. The two‐dimensional MSE of each intrinsic modal component is used as a learning sample for VRVM. Finally, the bearing fault data under 1 hp load are taken as training samples, and the bearing fault data under two loads of 0 hp and 3 hp are used as test samples to verify the effectiveness of the intelligent diagnosis model. The experimental results show that the intelligent fault diagnosis method proposed in this paper can accurately diagnose the type of fault and the degree of damage and has high robustness.

An Extreme-Learning-Machine-Based Hyperspectral Detection Method of Insulator Pollution Degree

The on-line detection of insulator pollution degree of transmission lines is important to the prevention and control of flashover. This paper proposed a non-contact detection method of insulator pollution degree based on hyperspectral technique. Firstly, hyperspectral images of the samples with different pollution degrees were obtained by hyper-spectrometer. Secondly, after original hyperspectral images were corrected by black-and-white correction and multiplicative scatter correction, hyperspectral curves from the region of interest (ROI) of corrected images were obtained. Finally, a multiclassification model of extreme learning machine (ELM) was built to realize the pollution degree classification of test samples. The results show that the absorption peak, the position of reflection peak, amplitude and the change trend of the hyperspectral curve obviously change with different kinds of pollution on the surface of silicone rubber, whereas only the amplitude obviously changes with same kind of pollution on the surface of silicone rubber; and the ELM-classification model can accurately and rapidly classify the pollution degree, with the pollution degree classification accuracy of NaCl, CaSO4 and mixed NaCl-CaSO4 respectively reaching 95%, 97.5% and 97.5%; and finally The ELM model based on hyperspectral curves data of the artificial pollution samples can classify the surface of insulator umbrellas with different pollution degrees, and the classification accuracy of CaSO4 and mixed NaCl-CaSO4 samples respectively are 87.5% and 90%. Consequently, the results of this study prove that hyperspectral technique has considerable potential for the non-contact detection of insulator pollution degree.

Analysis of happiness in EU countries using the multi-model classification based on models of symbolic data

The results of happiness analysis are presented in the form of a World Happiness Report that covers 156 countries and 17 different indicators. In the article model-based clustering ensemble is built to determine what selected European countries have similar patterns of happiness. The results are analyzed using multidimensional scaling and a decision tree to find out what factors determine cluster memberships. In the empirical part, three clusters were detected The first contains countries: Austria, Denmark, Finland, Germany, Ireland, Luxembourg, the Netherlands, Norway, Sweden, Switzerland and the United Kingdom. They have the highest values for all the variables, except the negative affect. The second cluster contains seven countries: Bulgaria, Estonia, Hungary, Lithuania, Poland, Romania and Slovakia. This cluster is also the most homogeneous one. The third cluster contains eight countries: Cyprus, the Czech Republic, France, Greece, Italy, Portugal, Slovenia and Spain

DRFMM: a map-matching algorithm based on distributed random forest multi-classification

In the vehicle navigation system, the vehicle movement trajectory displayed on the electronic map reflects the results of real-time positioning by the GPS measuring device. Map matching is the process of matching a series of GPS coordinates to an electronic map to find the true path of the trajectory. In this study, a distributed random forest map-matching algorithm (DRFMM) is proposed. The point-line matching method is used as the basic feature. The random forest algorithm in the Spark platform is used to train the historical data. The road network data is meshed, and a multi-classification model is trained offline for each grid in the road network to predict the FCD data online. The experimental results show that the DRFMM algorithm proposed in this paper has improved the accuracy of point-line matching by 10%. The multi-classification method keeps the matching accuracy with the increase of data volume. At the same time, with multi-threading and distributed platform, DRFMM's matching speed is nearly 6 times faster than stand-alone matching algorithm.

Gradient boosting trees method in the task of software identification

Subject of Research.The paper proposes an approach to the use of gradient boosted decision trees algorithm. For this purpose, CatBoost algorithm developed by Yandex is proposed. Its implementation is aimed at the problem solution of OS Linux software identification in order to reduce the number of system vulnerabilities, which occur due to the installation of unauthorized software by automated system users. We consider an approach to the program signatures formation and further training of CatBoostClassifier classifier model. The subsequent recognition task is set for the identified programs that were not previously involved in the model training process. Method. Free CatBoost software was used for implementation of the gradient boosted decision trees algorithm. CatBoostClassifier multi-classification model was created on its basis. The use of this model allows identifying test sample elf-files.Main Results. The training parameters of the classification model are selected. An experiment is carried out to identify elf-files with the use of ten different featuresof emerging signature programs. The results obtained in the new approach are compared with the results of the previously developed method of identification based on the application of the statistical criterion of Chi-square homogeneity at the significance level p = 0.01. Practical Relevance. The results of the study can be recommended to information security specialists for data media audit. The developed approach gives the possibility to identify violations of the established security policy in the processing of confidential information.