Classification Accuracy Of Network Research Articles

Multi-label arrhythmia classification using 12-lead ECG based on lead feature guide network

BackgroundMulti-label arrhythmia classification plays a crucial role in the prevention and diagnosis of cardiac diseases. Intelligent analysis of multi-lead electrocardiogram (ECG) signals is the main method of classification. Existing methods have shown that different lead ECG signals contribute differently to the classification network. MethodsIn this paper, we propose a Lead Feature Guide Network (LFG-Net) to improve the feature representation capability of the network by focusing on the lead features with high contribution. Firstly, the lead features with high contributions are selected by the Shapley Additive exPlanations (SHAP) method. Secondly, the lead features are utilized to guide the classification model in bridging the gap between the selected features and the classification model by minimizing the difference in their losses. Finally, the network improves the accuracy of arrhythmia classification due to the extraction of more valid feature information. ConclusionThe results demonstrated that the F1-score reached 0.842 for classification on the China Physiological Signal Challenge (CPSC) 2018 dataset, which indicated the network outperformed other state-of-the-art methods. The parameters of the network during prediction are 1.02 M. Through experiments, it can be verified that the effectiveness of the proposed method on arrhythmia classification with multi-label, and it can also be demonstrated that the lead features with high contributions can improve the network classification accuracy.

Data Mining with Supervised Instance Selection Improves Artificial Neural Network Classification Accuracy

Data Mining with Supervised Instance Selection Improves Artificial Neural Network Classification Accuracy

Transfer learning evaluation based on optimal convolution neural networks architecture for bearing fault diagnosis applications

Intelligent fault diagnosis utilizing deep learning algorithms is currently a topic of great interest. When developing a new Convolutional Neural Network (CNN) architecture to address machine diagnosis problem, it is common to use a deep model, with many layers, many feature maps, and large kernels. These models are capable of learning complex relationships and can potentially achieve superior performance on test data. However, not only does a large network potentially impose undue computational complexity for training and eventual deployment, it may also lead to more brittleness—where data outside of the curated dataset used in CNN training and evaluation is poorly handled. Accordingly, this paper will investigate a methodical approach for identifying a quasi-optimal CNN architecture to maximize robustness when a model is trained under one set of operating conditions, and deployed under a different set of conditions. Optuna software will be used to optimize a baseline CNN model for robustness to different rotational speeds and bearing Model #’s. To further improve the network generalization capabilities, this paper proposes the addition of white Gaussian noise to the raw vibration training data. Results indicate that the number of trainable weights and associated multiplications in the optimized model were reduced by almost 95% without jeopardizing the network classification accuracy. Additionally, moderate Additive White Gaussian Noise (AWGN) improved the model adaptation capabilities.

Lightweight and highly robust memristor-based hybrid neural networks for electroencephalogram signal processing

Memristor-based neuromorphic computing shows great potential for high-speed and high-throughput signal processing applications, such as electroencephalogram (EEG) signal processing. Nonetheless, the size of one-transistor one-resistor (1T1R) memristor arrays is limited by the non-ideality of the devices, which prevents the hardware implementation of large and complex networks. In this work, we propose the depthwise separable convolution and bidirectional gate recurrent unit (DSC-BiGRU) network, a lightweight and highly robust hybrid neural network based on 1T1R arrays that enables efficient processing of EEG signals in the temporal, frequency and spatial domains by hybridizing DSC and BiGRU blocks. The network size is reduced and the network robustness is improved while ensuring the network classification accuracy. In the simulation, the measured non-idealities of the 1T1R array are brought into the network through statistical analysis. Compared with traditional convolutional networks, the network parameters are reduced by 95% and the network classification accuracy is improved by 21% at a 95% array yield rate and 5% tolerable error. This work demonstrates that lightweight and highly robust networks based on memristor arrays hold great promise for applications that rely on low consumption and high efficiency.

Automatic Modulation Classification with Neural Networks via Knowledge Distillation

Deep learning is used for automatic modulation recognition in neural networks, and because of the need for high classification accuracy, deeper and deeper networks are used. However, these are computationally very expensive for neural network training and inference, so its utility in the case of a mobile with memory limitations or weak computational power is questionable. As a result, a trade-off between network depth and network classification accuracy must be considered. To address this issue, we used a knowledge distillation method in this study to improve the classification accuracy of a small network model. First, we trained Inception–Resnet as a teacher network, which has a size of 311.77 MB and a final peak classification accuracy of 93.09%. We used the method to train convolutional neural network 3 (CNN3) and increase its peak classification accuracy from 79.81 to 89.36%, with a network size of 0.37 MB. It was also used similarly to train mini Inception–Resnet and increase its peak accuracy from 84.18 to 93.59%, with a network size of 39.69 MB. When we compared all classification accuracy peaks, we discover that knowledge distillation improved small networks and that the student network had the potential to outperform the teacher network. Using knowledge distillation, a small network model can achieve the classification accuracy of a large network model. In practice, choosing the appropriate student network based on the constraints of the usage conditions while using knowledge distillation (KD) would be a way to meet practical needs.

Time series data augmentation method of small sample based on optimized generative adversarial network

SummaryIn the time series classification task, for the problem that small samples have difficulty in capturing sequence features cause deep learning models have low classification accuracy. This article proposes a time series data augmentation method based on gated recurrent unit (GRU)‐convolutional neural network (CNN) structure of generative adversarial network, denoted as GC‐GAN. Specifically, the structure of GRU network in series with CNN is used as the generator of GAN model, while the discriminator part uses two‐layer convolution network. In order to avoid the disappearance of gradient, Wasserstein distance is used to measure the distance between the generated distribution and the true distribution, and penalties are introduced in the loss function. Finally, the particle swarm algorithm is used to optimize the hidden variable input of the generator to approximate the real sequence to the greatest extent and obtain the best sample. The experimental results on the UCR time series dataset show that the long short‐term memory network classification accuracy has been effectively improved after using the GC‐GAN model, with a maximum improvement of 4.4%, indicating that the method proposed in this article has better data augmentation capabilities.

Bit-Aware Fault-Tolerant Hybrid Retraining and Remapping Schemes for RRAM-Based Computing-in-Memory Systems

Non-volatile computing-in-memory (nvCIM) has become one of the most efficient methods to deal with increasingly complicated neural networks compared to the traditional von Neumann architecture. However, due to the immature fabrication issues, the yield of resistive random access memory (RRAM) cells is limited, which causes the Stuck-At-Faults (SAFs) and decrease of network classification accuracy. In this brief, the bit-aware fault-tolerant hybrid retraining and remapping schemes are proposed to restore the network classification accuracy caused by SAFs. The low weight bits are retrained with the fault mask and the high weight bits are remapped with column swap and zero correction methods. For those bits whose values are still wrong after remapping, the fault index is used to record and calibrate. Experimental results show that the classification accuracy of MNIST, CIFAR-10, and CIFAR-100 is restored from 12.54% to 98.13%, from 10.29% to 90.60%, and from 0.94% to 74.30% with only 10.3% energy consumption and 5.88% area overhead, respectively.

Supervised Computer-Aided Diagnosis (CAD) Methods for Classifying Alzheimer's Disease-Based Neurodegenerative Disorders.

Alzheimer's disease is incurable at the moment. If it can be appropriately diagnosed, the correct treatment can postpone the patient's illness. To aid in the diagnosis of Alzheimer's disease and to minimize the time and expense associated with manual diagnosis, a machine learning technique is employed, and a transfer learning method based on 3D MRI data is proposed. Machine learning algorithms can dramatically reduce the time and effort required for human treatment of Alzheimer's disease. This approach extracts bottleneck features from the M-Net migration network and then adds a top layer to supervised training to further decrease the dimensionality and delete portions. As a consequence, the transfer network presented in this study has several advantages in terms of computational efficiency and training time savings when used as a machine learning approach for AD-assisted diagnosis. Finally, the properties of all subject slices are combined and trained in the classification layer, completing the categorization of Alzheimer's disease symptoms and standard control. The results show that this strategy has a 1.5 percentage point better classification accuracy than the one that relies exclusively on VGG16 to extract bottleneck features. This strategy could cut the time it takes for the network to learn and improve its ability to classify things. The experiment shows that the method works by using data from OASIS. A typical transfer learning network's classification accuracy is about 8% better with this method than with a typical network, and it takes about 1/60 of the time with this method.

Structured Network Pruning via Adversarial Multi-indicator Architecture Selection

Network pruning offers an opportunity to facilitate deploying convolutional neural networks (CNNs) on resource-limited embedded devices. Pruning more redundant network structures while ensuring network accuracy is challenging. Most existing CNN compression methods iteratively prune the “least important” filters and retrain the pruned network layer-by-layer, which may lead to a sub-optimal solution. In this paper, an end-to-end structured network pruning method based on adversarial multi-indicator architecture selection (AMAS) is presented. The pruning is implemented by striving to align the output of the baseline network with the output of the pruned network in a generative adversarial framework. Furthermore, to efficiently find optimal pruned architecture under constrained resources, an adversarial fine-tuning network selection strategy is designed, in which two contradictory indicators, namely pruned channel number and network classification accuracy, are considered. Experiments on SVHN show that AMAS reduces 75.37% of FLOPs and 74.42% of parameters with even 0.36% accuracy improvement for ResNet-110. On CIFAR-10, it achieves a reduction of 77.08% FLOPs and removes 73.98% of parameters with negligible accuracy cost for GoogLeNet. In particular, it obtains a 56.87% pruned rate in FLOPs and 59.18% parameters reduction, while with an increase of 0.49% accuracy for ResNet-110, which significantly outperforms state-of-the-art methods.

Behavior analysis method for indoor environment based on app usage mining

With the recent development of information and communication technologies, the utilization areas of spatial information are increasing rapidly while merging with various industries and technologies. This study introduces a system that can analyze the space utilization of users at low cost in indoor environment using simple smartphone app logs. We collect and process important information from mobile app logs and Google app server and generate a high-dimensional dataset required to analyze user behaviors. In addition, user behaviors are classified and clustered by applying a VGG classifier and a clustering algorithm based on t-stochastic neighbor embedding (t-SNE). Our system can easily acquire a large amount of data required for deep learning network learning without additional sensors for spatial analysis and enhance the accuracy of network classification and cluster through these data. Our methodology can assist spatial analysis for indoor environments in which people are living and help reduce the cost of space utilization feedback from users.

Cardiac magnetic resonance image diagnosis of hypertrophic obstructive cardiomyopathy based on a double-branch neural network

Objective: Cardiac magnetic resonance (CMR) imaging is a well-established technique for diagnosis of hypertrophic obstructive cardiomyopathy (HOCM) and evaluation of cardiac function, but the process is complicated and time consuming. Therefore, this paper proposes a cardiomyopathy recognition algorithm using a multi-task learning mechanism and a double-branch deep learning neural network.Method: We implemented a double-branch neural network CMR-based HOCM recognition algorithm. Compared with the traditional classification algorithms such as the ResNet, DenseNet network, contrast the accuracy of network classification of cardiomyopathy is higher by 10.11%.Result: The loss curve of the algorithm basically converges in 100 rounds, and the convergence speed of the algorithm is twice that of the traditional algorithm. The accuracy of this algorithm to classify cardiomyopathy is 96.79%, and the sensitivity is 95.24%, which is 10.11% higher than the conventional algorithm.Conclusion: The CMR imaging automatic recognition algorithm for HOCM capture static morphological and motion characteristics of the heart, and comprehensively enhances recognition accuracy when the sample size is limited.

Extracting Classification Rules from Artificial Neural Network Trained with Discretized Inputs

Rule extraction from artificial neural networks remains important task in complex diseases such as diabetes and breast cancer where the rules should be accurate and comprehensible. The quality of rules is improved by the improvement of the network classification accuracy which is done by the discretization of input attributes. In this paper, we developed a rule extraction algorithm based on multiobjective genetic algorithms and association rules mining to extract highly accurate and comprehensible classification rules from ANN’s that have been trained using the discretization of the continuous attributes. The data pre-processing provides very good improvement of the ANN accuracy and consequently leads to improve the performance of the classification rules in terms of fidelity and coverage. The results show that our algorithm is very suitable for medical decision making, so an excellent average accuracy of 94.73 has been achieved for the Pima dataset and 99.36 for the breast cancer dataset.

Sanders classification of calcaneal fractures in CT images with deep learning and differential data augmentation techniques

BackgroundClassification of the type of calcaneal fracture on CT images is essential in driving treatment. However, human-based classification can be challenging due to anatomical complexities and CT image constraints. The use of computer-aided classification system in standard practice is additionally hindered by the availability of training images. The aims of this study is to 1) propose a deep learning network combined with data augmentation technique to classify calcaneal fractures on CT images into the Sanders system, and 2) assess the efficiency of such approach with differential training methods. MethodsIn this study, the Principle component analysis (PCA) network was selected for the deep learning neural network architecture for its superior performance. CT calcaneal images were processed through PCA filters, binary hashing, and a block-wise histogram. The Augmentor pipeline including rotation, distortion, and flips was applied to generate artificial calcaneus fractured images. Two types of training approaches and five data sample sizes were investigated to evaluate the performance of the proposed system with and without data augmentation. ResultsCompared to the original performance, use of augmented images during training improved network performance accuracy by almost twofold in classifying Sanders fracture types for all dataset sizes. A fivefold increase in the number of augmented training images improved network classification accuracy by 35%. The proposed deep CNN model achieved 72% accuracy in classifying CT calcaneal images into the four Sanders categories when trained with sufficient augmented artificial images. ConclusionThe proposed deep-learning algorithm coupled with data augmentation provides a feasible and efficient approach to the use of computer-aided system in assisting physicians in evaluating calcaneal fracture types.

Voltage-Controlled Spintronic Stochastic Neuron for Restricted Boltzmann Machine With Weight Sparsity

This work proposes a novel three-terminal magnetic tunnel junction (MTJ) as a stochastic neuron. The neuron is probabilistically switched based on the voltage-controlled magnetic anisotropy (VCMA) effect with the assistance of Rashba effective field. We find that a restricted Boltzmann machine (RBM) implemented using our proposed neuron for handwritten character recognition can achieve synaptic weight sparsity, without sacrificing the network classification accuracy. Moreover, the RBM implemented by this novel neuron performs even better in the presence of device variations, implying that our device is highly suitable for the hardware implementation of RBM.

A systematic approach for segmenting voiced/unvoiced signals using fuzzy-logic system and general fusion of neural network models for phonemes-based speech recognition

In this paper, a speech-to-text translation model has been developed for Malaysian speakers based on 41 classes of Phonemes. A simple data acquisition algorithm has been used to develop a MATLAB graphical user interface (GUI) for recording the isolated word speech signals from 35 non-native Malaysian speakers. The collected database consists of 86 words with 41 classes of phoneme based on Affricatives, Diphthongs, Fricatives, Liquid, Nasals, Semivowels and Glides, Stop and Vowels. The speech samples are preprocessed to eliminate the undesirable artifacts and the fuzzy voice classifier has been employed to classify the samples into voiced sequence and unvoiced sequence. The voiced sequences are divided into frame segments and for each frame, the Linear Predictive co-efficients features are obtained from the voiced sequence. Then the feature sets are formed by deriving the LPC features from all the extracted voiced sequences, and used for classification. The isolated words chosen based on the phonemes are associated with the extracted features to establish classification system input-output mapping. The data are then normalized and randomized to rearrange the values into definite range. The Multilayer Neural Network (MLNN) model has been developed with four combinations of input and hidden activation functions. The neural network models are trained with 60%, 70% and 80% of the total data samples. The neural network architecture was aimed at creating a robust model with 60%, 70%, and 80% of the feature set with 25 trials. The trained network model is validated by simulating the network with the remaining 40%, 30%, and 20% of the set. The reliability of trained network models were compared by measuring true-positive, false-negative, and network classification accuracy. The LPC features show better discrimination and the MLNN neural network models trained using the LPC spectral band features gives better recognition.

Automatic Classification of Anterior Chamber Angle Using Ultrasound Biomicroscopy and Deep Learning.

PurposeTo develop a software package for automated classification of anterior chamber angle of the eye by using ultrasound biomicroscopy.MethodsUltrasound biomicroscopy images were collected, and the trabecular-iris angle was manually measured and classified into three categories: open angle, narrow angle, and angle closure. Inception v3 was used as the classifying convolutional neural network and the algorithm was trained.ResultsWith a recall rate of 97% in the test set, the neural network's classification accuracy can reach 97.2% and the overall area under the curve was 0.988. The sensitivity and specificity were 98.04% and 99.09% for the open angle, 96.30% and 98.13% for the narrow angle, and 98.21% and 99.05% for the angle closure categories, respectively.ConclusionsPreliminary results show that an automated classification of the anterior chamber angle achieved satisfying sensitivity and specificity and could be helpful in clinical practice.Translational RelevanceThe present work suggests that the algorithm described here could be useful in the categorizing of anterior chamber angle and screening for subjects who are at high risk of angle closure.

Application of Hyperspectral Image Classification Based on Overlap Pooling

Convolutional neural networks (CNN) are increasingly being used in hyperspectral image (HSI) classification. However, most pooling methods are non-overlap pooling and ignore the influence of neighboring pixels on image characteristics, thereby limiting network classification accuracy. This work presents a deep CNN that is based on overlap pooling; in this model, non-overlap pooling is replaced with overlap pooling to improve the accuracy of feature extraction. However, overlap pooling introduces additional noise while improving feature accuracy. We have found that different combinations of max pooling and mean pooling can effectively solve the problem and significantly improve classification performance. The best pooling combination (max–mean–mean) for HSI classification is obtained after verification through experiments. A rectified linear unit activation function layer and the softmax loss classification model are combined to improve overall classification accuracy. Experiments on three HSI data sets, namely, Indian Pines, Salinas and Pavia University, show that the CNN model can increase overall accuracy to 95.66, 97.8 and 97.48%, respectively. Compared with deep network models such as deep belief network and non-overlap CNN, the proposed model has significantly improved the classification accuracy, and thus verifying the high accuracy of feature extraction of overlap pooling in CNN.

Klasifikasi Paket Jaringan Berbasis Analisis Statistik dan Neural Network

Distributed Denial-of-Service (DDoS) is one of network attack technique which increased every year, especially in both of intensity and volume. DDoS attacks are still one of the world's major Internet threats and become a major problem of cyber-world security. Research in this paper aims to establish a new approach on network packets classification, which can be a basis for framework development on Distributed Denial-of-Service (DDoS) attack detection systems. The proposed approach to solving the problem on network packet classification is by combining statistical data quantification methods with neural network methods. Based on the test, it is found that the average percentage of neural network classification accuracy against network data packet is 92.99%.

Provide a data mining algorithm for text classification based on text content emotions using neural network

One of the newest areas of research is in data mining and text mining is automatic discovery of knowledge from semi-structured text. an important application of data mining in the classification texts.neural networks have emerged as a powerful tool in the classification of the content of texts and a promising alternative to conventional methods of classification. in this study, by combining and improving the regulatory procedure parameters weighted binary artificial neural network model, we will provide an algorithm to classify content. This algorithm content of the texts of the Persian texts will be based on the polarity of emotion suitable performance and high accuracy. The proposed algorithm is implemented using the simulation software MATLAB and evaluated collected over 1200 comments in Farsi in the real environment. The results above show that the proposed algorithm is a neural network classification accuracy of 96% negative polarity and positive polarity sentences based on document content.

基于主成分分析和BP神经网络的柑橘黄龙病诊断技术

To address the limitations of conventional techniques,a method of principal component analysis and BP neural network was discussed to diagnose and classify citrus HuangLongBing.Data was obtained by a hyperspectral imaging system with the wavelength range of 370~988nm,its high dimension data was reduced by principal component analysis,and then BP neural network was used to model for classification.The results showed that the first four principal components cumulative variance contribution rate achieved 97.42%.On one hand,BP neural network classification accuracy rate achieved 85% or more;on the other hand,after the principal component analysis,classification of BP neural network accuracy substantially was more than 90%.This method for nondestructive testing of citrus HuangLongBing is feasible.