Data analysis in relativistic nuclear physics using neural network

Two neural network algorithms for data analysis in relativistic nuclear physics are presented. A neural network technique (Hopfield method) is used in order to reconstruct particle tracks starting from a data set obtained with a coordinate detector system. An algorithm for circles recognition using deformable templates is carried out and its performances are studied. The technical limitations of the detectors, which in real situation prevent the possibility to reconstruct hits right on the circle, and presence of the noise points are taken into account.     

Visibility graph analysis on quarterly macroeconomic series of China based on complex network theory

The visibility graph approach and complex network theory provide a new insight into time series analysis. The inheritance of the visibility graph from the original time series was further explored in the paper. We found that degree distributions of visibility graphs extracted from Pseudo Brownian Motion series obtained by the Frequency Domain algorithm exhibit exponential behaviors, in which the exponential exponent is a binomial function of the Hurst index inherited in the time series. Our simulations presented that the quantitative relations between the Hurst indexes and the exponents of degree distribution function are different for different series and the visibility graph inherits some important features of the original time series. Further, we convert some quarterly macroeconomic series including the growth rates of value-added of three industry series and the growth rates of Gross Domestic Product series of China to graphs by the visibility algorithm and explore the topological properties of graphs associated from the four macroeconomic series, namely, the degree distribution and correlations, the clustering coefficient, the average path length, and community structure. Based on complex network analysis we find degree distributions of associated networks from the growth rates of value-added of three industry series are almost exponential and the degree distributions of associated networks from the growth rates of GDP series are scale free. We also discussed the assortativity and disassortativity of the four associated networks as they are related to the evolutionary process of the original macroeconomic series. All the constructed networks have “small-world” features. The community structures of associated networks suggest dynamic changes of the original macroeconomic series. We also detected the relationship among government policy changes, community structures of associated networks and macroeconomic dynamics. We find great influences of government policies in China on the changes of dynamics of GDP and the three industries adjustment. The work in our paper provides a new way to understand the dynamics of economic development.     

Vibrational analysis of carbon nanotubes using molecular mechanics and artificial neural network

This paper presents the molecular mechanics based finite element modeling of carbon nanotubes (CNTs) and their applications as mass sensors. The beam element with elastic behavior is considered as the bond between the carbon atoms and its properties are obtained using equating continuum and molecular characteristics. The first five natural frequencies of CNTs in cantilever and doubly clamped boundary conditions (BCs) and their corresponding mode shapes are studied in detail. Furthermore, a multilayer perceptron neural network is used to predict the fundamental vibration frequencies of the CNTs with different diameters and lengths. In addition, variations of the natural frequencies of the CNTs with distorted cross sections are investigated. Moreover, the effects of some attached masses with various values on the first three natural frequencies of a considered CNT are studied here.     

A 3D densely connected convolution neural network with connection-wise attention mechanism for Alzheimer's disease classification

PurposeAlzheimer's disease (AD) is a progressive and irreversible neurodegenerative disease. In recent years, machine learning methods have been widely used on analysis of neuroimage for quantitative evaluation and computer-aided diagnosis of AD or prediction on the conversion from mild cognitive impairment (MCI) to AD. In this study, we aimed to develop a new deep learning method to detect or predict AD in an efficient way.Materials and methodsWe proposed a densely connected convolution neural network with connection-wise attention mechanism to learn the multi-level features of brain MR images for AD classification. We used the densely connected neural network to extract multi-scale features from pre-processed images, and connection-wise attention mechanism was applied to combine connections among features from different layers to hierarchically transform the MR images into more compact high-level features. Furthermore, we extended the convolution operation to 3D to capture the spatial information of MRI. The features extracted from each 3D convolution layer were integrated with features from all preceding layers with different attention, and were finally used for classification. Our method was evaluated on the baseline MRI of 968 subjects from ADNI database to discriminate (1) AD versus healthy subjects, (2) MCI converters versus healthy subjects, and (3) MCI converters versus non-converters.ResultsThe proposed method achieved 97.35% accuracy for distinguishing AD patients from healthy control, 87.82% for MCI converters against healthy control, and 78.79% for MCI converters against non-converters. Compared with some neural networks and methods reported in recent studies, the classification performance of our proposed algorithm was among the top ranks and improved in discriminating MCI subjects who were in high risks of conversion to AD.ConclusionsDeep learning techniques provide a powerful tool to explore minute but intricate characteristics in MR images which may facilitate early diagnosis and prediction of AD.     

Sensitivity analysis for estimation of power losses in magnetic cores using neural network

Experimental data from a sample of 42 cores made from grain oriented 0.27 mm thick 3% SiFe electrical steel with dimensions ranging from 35 to 160 mm outer diameter, 25-100 mm inner diameter and 10-70 mm strip width and a flux density range 0.2-1.7 T have been obtained at 50 Hz and used as training data to a feed forward neural network. An analytical equation for prediction of power loss as depends on input parameters from the results of sensitivity analysis has been obtained. The calculated power losses with the analytical expression have also been compared with power loss obtained from the Preisach model after it has been applied to toroidal cores. The results show the proposed model can be used for estimation of power losses in the toroidal cores.     

Classifying cosmic-ray proton and light groups in LHAASO-KM2A experiment with graph neural network

The precise measurement of cosmic-ray(CR) knees of different primaries is essential to reveal CR acceleration and propagation mechanisms, as well as to explore new physics. However, the classification of CR components is a difficult task, especially for groups with similar atomic numbers. Given that deep learning achieved remarkable breakthroughs in numerous fields, we seek to leverage this technology to improve the classification performance of the CR Proton and Light groups in the LHAASO-KM2A experiment. In this study, we propose a fused graph neural network model for KM2A arrays, where the activated detectors are structured into graphs. We find that the signal and background are effectively discriminated in this model, and its performance outperforms both the traditional physicsbased method and the convolutional neural network(CNN)-based model across the entire energy range.     

基于注意力机制与图卷积神经网络的单目红外图像深度估计

对场景中的物体进行深度估计是无人驾驶领域中的关键问题,红外图像有利于在光线不佳的情况下解决深度估计问题.针对红外图像纹理不清晰与边缘信息不丰富的特点,提出了将注意力机制与图卷积神经网络相结合来解决单目红外图像深度估计问题.首先,在深度估计问题中,图像中每个像素点的深度信息不仅与其周围像素点的深度信息相关,还需考虑更大范...     

Temporal and dimensional effects in evolutionary graph theory

The spread in time of a mutation through a population is studied analytically and computationally in fully connected networks and on spatial lattices. The time t* for a favorable mutation to dominate scales with the population size N as N(D+1)/D in D-dimensional hypercubic lattices and as NlnN in fully-connected graphs. It is shown that the surface of the interface between mutants and nonmutants is crucial in predicting the dynamics of the system. Network topology has a significant effect on the equilibrium fitness of a simple population model incorporating multiple mutations and sexual reproduction.     

超声探伤信号的人工神经网络识别

粗晶奥氏体不锈钢的超声探伤受到能否有效区分有用信号与背影噪声的限制，目前人们大多倾向使用频率分隔来提高缺陷回波比例。本文则介绍一种用傅里叶变换作特征提取，用前馈网络自动识别奥氏体钢中缺陷信号的方法。在作者的实验中，这种方法的正确识别率达到９０％。     

Efficient detection of refugees and migrants in Turkey using convolutional neural network

At the start of the Syrian Civil War in 2011, NGOs played a big part in giving refugees access to aid and distributing that aid so that people could go to school, get a job, or get medical care. Within the last few years, when tensions rose between Syrian refugees and the Turkish community, many non-governmental organizations switched their attention to fostering community among refugees in Turkey. Over the past two decades, family displacement has become a big problem in various countries due to a rise in the frequency with which natural catastrophes, military conflicts, and terrorist strikes occur. It poses severe difficulties for governing bodies and the organizations that oversee them. This research aims to identify and track refugees in surveillance zones by utilizing artificial intelligence. Refugees are vulnerable to acts of nature and human aggression, which makes their random relocation or encampments challenging to manage. To overcome these challenges, a convolutional neural network deep learning model has been proposed to identify and track refugees in surveillance zones. The proposed solution is integrated with Internet of Things (IoT) technology by equipping the system with IoT sensors to capture real-time data on the location and movements of refugees. This combination of AI and IoT has the potential to improve the efficiency and effectiveness of refugee management efforts. The suggested solution uses a convolutional neural network deep learning model, which can quickly identify a refugee’s face. To assist the government in locating a specific refugee, the system simultaneously connects with the refugees and requests that they regularly update their location. The system alerts security to identify the missing immigrant since the refugee does not update their whereabouts. Without human intervention, the deep learning algorithm makes it simple to recognize immigrants and keep an eye on them.     

利用几何特性及神经网络进行人脸探测技术的研究

在人脸识别过程中 ,首先也是最重要的一个环节是人脸探测 ,因为一旦从图像中定位并提取到了人脸 ,那么下一步的人脸识别工作就变得非常容易。眼睛是人脸图像中最容易探测的部位 ,而且通过探测双眼来发现人脸最符合人的视觉习惯。提出了一种基于几何特征分析和人工神经网络的由粗到细的两级人脸探测方法。在第一级中 ,眼睛和脸是通过测量眼睛的尺寸和眼睛与脸的位置关系探测到的 ,第一级的输出是一个尺寸归一化的人脸 ,但偶尔也伴随着一个或多个因对复杂背景中与眼睛类似的物体的误判而得到的非人脸图像 ;第二级神经网络正是用来过滤掉第一级中被误判的人脸。实验表明 ,这种由粗到细的两级人脸探测系统具有很高的稳定性和探测正确率     

卷积神经网络的缺陷类型识别分析

该文提出一种基于卷积神经网络直接对阵列超声检测原始信号进行缺陷类型识别的方法,该方法无需对超声回波原始信号进行特征提取.文章研究对比了不同卷积神经网络及其优化的识别性能.首先采用超声相控阵系统对不同试块上的平底孔、球底孔、通孔三种缺陷进行超声检测,然后利用LeNet5、VGG16和ResNet三种卷积神经网络对一维和二...     

A comparison in a back-bead prediction of gas metal arc welding using multiple regression analysis and artificial neural network

This research was done on the basis of prediction that there is a relationship between welding parameters and geometry of the back-bead in arc welding which is a gap. Multiple regression analysis and artificial neural network were used as methods for predicting the geometry of the back-bead. The multiple regression analysis and the artificial neural network were formed, and the analysis data or verification data which were used in the formation process of the multiple regression, and the training data or test data which were used in the formation process of the artificial neural network, were used to perform the prediction of the back-bead. Through this research, it was found that the error rate predicted by the artificial neural network was smaller than that predicted by the multiple regression analysis, in terms of the width and depth of the back-bead. It was also found that between the two predictions, the prediction of the width of the back-bead was superior to the prediction of the depth in both methods.     

Image noise cancellation using linear matrix inequality and cellular neural network

In this paper, the technique of image noise cancellation is presented by employing cellular neural networks (CNN) and linear matrix inequality (LMI). The main objective is to obtain the templates of CNN by using a corrupted image and a corresponding desired image. A criterion for the uniqueness and global asymptotic stability of the equilibrium point of CNN is presented based on the Lyapunov stability theorem (i.e., the feedback template “A” of CNN is solved at this step), and the input template “B” of CNN is designed to achieve desirable output by using the property of saturation nonlinearity of CNN. It is shown that the problem of image noise cancellation can be characterized in terms of LMIs. The simulation results indicate that the proposed method is useful for practical application.     

In vitro microemboli classification using neural network models and RF signals

Emboli classification is of high clinical importance for selecting appropriate treatment for patients. Several ultrasonic (US) methods using Doppler processing have been used for emboli detection and classification as solid or gaseous matter. We suggest in this experimental study exploiting the Radio-Frequency (RF) signal backscattered by the emboli since they contain additional information on the embolus than the Doppler signal. The aim of the study is the analysis of RF signals using Multilayer Perceptron (MLP) and Radial-Basis Function Network (RBFN) in order to classify emboli.Anthares scanner with RF access was used with a transmit frequency of 1.82 MHz at two mechanical indices (MI) 0.2 and 0.6. The mechanical index is given as the peak negative pressure (in MPa) divided by the square root of the frequency (in MHz). A Doppler flow phantom was used containing a 0.8 mm diameter vessel surrounded by a tissue mimicking material. To imitate gas emboli US behaviour, Sonovue microbubbles were injected at two different doses (10μl and 5μl) in a nonrecirculating at a constant flow. The surrounding tissue was assumed to behave as a solid emboli. In order to mimic real clinical pathological situations, Sonovue concentration was chosen such that the fundamental scattering from the tissue and from the contrast were identical. The amplitudes and bandwidths of the fundamental and the 2nd harmonic components were selected as input parameters to the MLP and RBFN models. Moreover the frequency bandwidths of the fundamental and the 2nd harmonic echoes were approximated by Gaussian functions and the coefficients were used as a third input parameter to the neural network models. The results show that the Gaussian coefficients provide the highest rate of classification in comparison to the amplitudes and the bandwidths of the fundamental and the 2nd harmonic components. The classification rates reached 89.28% and 92.85% with MLP and RBFN models respectively.This short communication demonstrates the opportunity to classify emboli based on a RF signals and neural network analysis.     

采用上下文相关的注意力机制及循环神经网络的语音增强方法

提出了采用上下文相关的注意力机制及循环神经网络的语音增强方法。该方法在训练阶段联合训练计算注意力评分的多层感知机和增强语音的深度循环网络,在测试阶段计算每一帧语音的注意力向量并与该帧语音拼接输入深度循环网络增强。在不同信噪比的实验中,该方法相比基线模型能更好地提高语音质量和可懂度,-6 dB下相对带噪语音短时客观可懂度(STOI)和语音质量感知评估(PESQ)可分别提高0.16和0.77,同时在未知噪声条件下该方法性能仍最优或接近最优。因此注意力机制可以有效强化模型对上下文信息的利用能力,从而提高模型增强性能。     

Individual identification using multi-metric of DTI in Alzheimer's disease and mild cognitive impairment

Accurate identification of Alzheimer's disease(AD) and mild cognitive impairment(MCI) is crucial so as to improve diagnosis techniques and to better understand the neurodegenerative process. In this work, we aim to apply the machine learning method to individual identification and identify the discriminate features associated with AD and MCI. Diffusion tensor imaging scans of 48 patients with AD, 39 patients with late MCI, 75 patients with early MCI, and 51 age-matched healthy controls(HCs) are acquired from the Alzheimer's Disease Neuroimaging Initiative database. In addition to the common fractional anisotropy, mean diffusivity, axial diffusivity, and radial diffusivity metrics, there are two novel metrics,named local diffusion homogeneity that used Spearman's rank correlation coefficient and Kendall's coefficient concordance,which are taken as classification metrics. The recursive feature elimination method for support vector machine(SVM)and logistic regression(LR) combined with leave-one-out cross validation are applied to determine the optimal feature dimensions. Then the SVM and LR methods perform the classification process and compare the classification performance.The results show that not only can the multi-type combined metrics obtain higher accuracy than the single metric, but also the SVM classifier with multi-type combined metrics has better classification performance than the LR classifier.Statistically, the average accuracy of the combined metric is more than 92% for all between-group comparisons of SVM classifier. In addition to the high recognition rate, significant differences are found in the statistical analysis of cognitive scores between groups. We further execute the permutation test, receiver operating characteristic curves, and area under the curve to validate the robustness of the classifiers, and indicate that the SVM classifier is more stable and efficient than the LR classifier. Finally, the uncinated fasciculus, cingulum, corpus callosum, corona radiate, external capsule, and internal capsule have been regarded as the most important white matter tracts to identify AD, MCI, and HC. Our findings reveal a guidance role for machine-learning based image analysis on clinical diagnosis.     

水下高分辨率声图中小目标的深度网络分类方法

针对声成像数据缺少条件下的水下沉底小目标分类问题,提出一种深度网络分类算法。首先,采用高斯混合模型对声影区统计特性进行建模并提取声图阴影,在此基础上构建仿真数据集和真实数据集。将仿真数据集输入卷积神经网络进行训练,保留其特征提取部分,用于对真实数据集进行特征提取.重建网络分类部分并采用真实数据集的特征向量进行训练。结果表明,所提出的方法分类正确率可达88.24%,与6种对照方法相比平均分类正确率分别提升8.67%,20.47%,19.78%,11.59%,9.01%,11.58%。验证了所提出方法在小样本条件下具有较好对水下沉底小目标的分类能力。其学习曲线收敛到96.25%,仅比验证曲线高5.14%,说明在一定程度上缓解了过拟合问题。将改进的卷积神经网络应用于融合分类器,通过与逻辑回归分类器、支持向量机对目标进行分类并融合决策,正确率为93.33%,可进一步提高算法的正确率和稳定性.