共查询到20条相似文献，搜索用时 187 毫秒
1.
《Information processing & management》2022,59(3):102916
Graph convolutional network (GCN) is a powerful tool to process the graph data and has achieved satisfactory performance in the task of node classification. In general, GCN uses a fixed graph to guide the graph convolutional operation. However, the fixed graph from the original feature space may contain noises or outliers, which may degrade the effectiveness of GCN. To address this issue, in this paper, we propose a robust graph learning convolutional network (RGLCN). Specifically, we design a robust graph learning model based on the sparse constraint and strong connectivity constraint to achieve the smoothness of the graph learning. In addition, we introduce graph learning model into GCN to explore the representative information, aiming to learning a high-quality graph for the downstream task. Experiments on citation network datasets show that the proposed RGLCN outperforms the existing comparison methods with respect to the task of node classification. 相似文献
2.
[目的/意义]在非结构化语料集中抽取知识要素,是实现知识图谱的重要环节,本文探索了应用深度学习中的卷积神经网络（CNN）模型进行旅游领域知识关系抽取方法。[方法/过程]抓取专业旅游网站的相关数据建立语料库,对部分语料进行人工标注作为训练集和测试集,通过Python语言编程实现分词、向量化及CNN模型,进行关系抽取实验。[结果/结论]实验结果表明,应用卷积神经网络对非结构化的旅游文本进行关系抽取时能够取得满意的效果（Precision 0.77,Recall 0.76,F1-measure 0.76）。抽取结果通过人工校对进行优化后,可以为旅游知识图谱构建、领域本体构建等工作奠定基础。 相似文献
3.
《Information processing & management》2023,60(4):103350
Knowledge graphs are sizeable graph-structured knowledge with both abstract and concrete concepts in the form of entities and relations. Recently, convolutional neural networks have achieved outstanding results for more expressive representations of knowledge graphs. However, existing deep learning-based models exploit semantic information from single-level feature interaction, potentially limiting expressiveness. We propose a knowledge graph embedding model with an attention-based high-low level features interaction convolutional network called ConvHLE to alleviate this issue. This model effectively harvests richer semantic information and generates more expressive representations. Concretely, the multilayer convolutional neural network is utilized to fuse high-low level features. Then, features in fused feature maps interact with other informative neighbors through the criss-cross attention mechanism, which expands the receptive fields and boosts the quality of interactions. Finally, a plausibility score function is proposed for the evaluation of our model. The performance of ConvHLE is experimentally investigated on six benchmark datasets with individual characteristics. Extensive experimental results prove that ConvHLE learns more expressive and discriminative feature representations and has outperformed other state-of-the-art baselines over most metrics when addressing link prediction tasks. Comparing MRR and Hits@1 on FB15K-237, our model outperforms the baseline ConvE by 13.5% and 16.0%, respectively. 相似文献
4.
针对钢板表面缺陷图像分类传统深度学习算法中需要大量标签数据的问题,提出一种基于主动学习的高效分类方法。该方法包含一个轻量级的卷积神经网络和一个基于不确定性的主动学习样本筛选策略。神经网络采用简化的convolutional base进行特征提取,然后用全局池化层替换掉传统密集连接分类器中的隐藏层来减轻过拟合。为了更好的衡量模型对未标签图像样本所属类别的不确定性,首先将未标签图像样本传入到用标签图像样本训练好的模型,得到模型对每一个未标签样本关于标签的概率分布(probability distribution over classes, PDC),然后用此模型对标签样本进行预测并得到模型对每个标签的平均PDC。将两类分布的KL-divergence值作为不确定性指标来筛选未标签图像进行人工标注。根据在NEU-CLS开源缺陷数据集上的对比实验,该方法可以通过44%的标签数据实现97%的准确率,极大降低标注成本。 相似文献
5.
[目的/意义]对社交媒体中热门新闻的及时识别,有助于加速正面资讯的投送或抑制负面资讯的扩散。当前,基于自然语言处理的传统识别方法正面临社交媒体新生态的挑战：大量新闻内容以图片、音视频形式存在,缺乏用于语义及情感分析的文本。[方法/过程]对此,本文首先将社交网络划分为众多社群,并按其层次结构组织为贝叶斯网络。接着,面向社群构建基于卷积神经网络的热门新闻识别模型,模型综合考虑新闻传播的宏观统计规律及微观传播过程,以提取社群内热门新闻传播的特征。最后,利用贝叶斯推理并结合局部性的模型识别结果进行全局性热度预测。[结果/结论]实验表明,本方法在语义缺失场景下可有效识别热门新闻,其准确度强于基于语义信息的机器学习方法,模型具有良好的时效性、可扩展性和适用性。该研究有助于社交媒体的监管机构及时识别出各类不含语义信息且迅速扩散的热点内容。 相似文献
6.
In this paper, a modified adaptive neural network for the compensation of deadzone is described, and simulated on a hydraulic positioning system, in which the dynamic model is separated into a series of connection of a nonlinear (deadzone) subsystem and a linear plant. The proposed approach uses two neural networks. One is the radial basis function (RBF) neural network, which is used for identifying parameters of deadzone. Based on the penalty function used in optimization theory, a multi-objective cost function with constraint is adopted to provide the best deadzone approximation. The result is used to train the other neural network for the inverse compensation of deadzone. The RBF neural network also generates the parameters of the linear plant for the design of an adaptive controller. A convergence analysis for the network training process is also presented. 相似文献
7.
Augmented reality is very useful in medical education because of the problem of having body organs in a regular classroom. In this paper, we propose to apply augmented reality to improve the way of teaching in medical schools and institutes. We propose a novel convolutional neural network (CNN) for gesture recognition, which recognizes the human's gestures as a certain instruction. We use augmented reality technology for anatomy learning, which simulates the scenarios where students can learn Anatomy with HoloLens instead of rare specimens. We have used the mesh reconstruction to reconstruct the 3D specimens. A user interface featured augment reality has been designed which fits the common process of anatomy learning. To improve the interaction services, we have applied gestures as an input source and improve the accuracy of gestures recognition by an updated deep convolutional neural network. Our proposed learning method includes many separated train procedures using cloud computing. Each train model and its related inputs have been sent to our cloud and the results are returned to the server. The suggested cloud includes windows and android devices, which are able to install deep convolutional learning libraries. Compared with previous gesture recognition, our approach is not only more accurate but also has more potential for adding new gestures. Furthermore, we have shown that neural networks can be combined with augmented reality as a rising field, and the great potential of augmented reality and neural networks to be employed for medical learning and education systems. 相似文献
8.
针对信息挖掘中的网页自动分类问题,提出了一种基于向量空间模型和并联BP网络的分类方法。该网络由并行连接的多个子网络组成,每个子网络负责一类模式特征的提取,多个子网并行处理所有模式,将分类结果在总输出层表现出来。以因特网上旅游网页分类为例验证了该方法的有效性。 相似文献
9.
《Information processing & management》2023,60(2):103223
Aspect-based sentiment analysis aims to determine sentiment polarities toward specific aspect terms within the same sentence or document. Most recent studies adopted attention-based neural network models to implicitly connect aspect terms with context words. However, these studies were limited by insufficient interaction between aspect terms and opinion words, leading to poor performance on robustness test sets. In addition, we have found that robustness test sets create new sentences that interfere with the original information of a sentence, which often makes the text too long and leads to the problem of long-distance dependence. Simultaneously, these new sentences produce more non-target aspect terms, misleading the model because of the lack of relevant knowledge guidance. This study proposes a knowledge guided multi-granularity graph convolutional neural network (KMGCN) to solve these problems. The multi-granularity attention mechanism is designed to enhance the interaction between aspect terms and opinion words. To address the long-distance dependence, KMGCN uses a graph convolutional network that relies on a semantic map based on fine-tuning pre-trained models. In particular, KMGCN uses a mask mechanism guided by conceptual knowledge to encounter more aspect terms (including target and non-target aspect terms). Experiments are conducted on 12 SemEval-2014 variant benchmarking datasets, and the results demonstrated the effectiveness of the proposed framework. 相似文献
10.
企业的信用风险评级是金融领域的一个重要问题,采用BP神经网络来研究上市公司的信用风险评价问题。首先构建了上市公司信用评价的财务指标体系,然后根据3个不同的隐层结点,生成3种不同的神经网络模型。设计7种不同的学习一验证比例,选取了不同行业上市公司的财务数据,利用MATLAB中的神经网络工具箱编程进行实证分析在哪种模型和学习一验证比例下能够更好的对企业进行信用风险评价。 相似文献
11.
采用面向对象的程序设计方法（ＯＯＰ）,利用Ｃ＋＋中的类模板,建立了神经网络模型的抽象类Ｎｅｔｗｏｒｋ并提出了用Ｃ＋＋实现的神经网络通用程序,该程序具有良好的操作笥和可扩展性,并以其在非线性动态系统辨识的应用为例予以说明。 相似文献
12.
《Information processing & management》2022,59(6):103095
Modeling discussions on social networks is a challenging task, especially if we consider sensitive topics, such as politics or healthcare. However, the knowledge hidden in these debates helps to investigate trends and opinions and to identify the cohesion of users when they deal with a specific topic. To this end, we propose a general multilayer network approach to investigate discussions on a social network. In order to prove the validity of our model, we apply it on a Twitter dataset containing tweets concerning opinions on COVID-19 vaccines. We extract a set of relevant hashtags (i.e., gold-standard hashtags) for each line of thought (i.e., pro-vaxxer, neutral, and anti-vaxxer). Then, thanks to our multilayer network model, we figure out that the anti-vaxxers tend to have ego networks denser (+14.39%) and more cohesive (+64.2%) than the ones of pro-vaxxer, which leads to a higher number of interactions among anti-vaxxers than pro-vaxxers (+393.89%). Finally, we report a comparison between our approach and one based on single networks analysis. We prove the effectiveness of our model to extract influencers having ego networks with more nodes (+40.46%), edges (+39.36%), and interactions with their neighbors (+28.56%) with respect to the other approach. As a result, these influential users are much more important to analyze and can provide more valuable information. 相似文献
13.
14.
《Journal of The Franklin Institute》2023,360(12):8178-8199
In large-scale complex dynamical networks, it is significant to estimate the states of target nodes with only a part of measured nodes. Meanwhile, multilayer complex dynamical networks exist widely in society and engineering. Therefore, it has important theoretic meaning and practical value to study the state estimation of target nodes in multilayer complex dynamical networks with limited node measurements. In this paper, with the measurable state information of a portion of nodes in one layer in the multilayer complex dynamical network, the state estimation of target nodes in other layers is studied. First, we build the model of the multilayer complex dynamical network which includes some target nodes and sensor nodes. Second, auxiliary nodes are selected by using the maximum matching principle in graph theory to construct the augmented node set. Third, we discuss the relationship between the minimum number of auxiliary nodes and interlayer connection probability in the multilayer complex dynamical network. Forth, an appropriate functional state observer is designed with limited number of measured nodes according to a typical model-based algorithm. Finally, numerical simulations are given to demonstrate the accuracy of the proposed method. The proposed method can achieve the accurate estimation with less placement of observers and fewer computational costs in the multilayer complex dynamical network. 相似文献
15.
16.
基于小波网络的电力系统短期负荷预报研究 总被引：7，自引：0，他引：7
本文结合小波和神经网络方法进行电力系统短期负荷预测的通用模型和方法的研究,建立了负荷预报的小波网络模型,确定了有效的算法求解小波函数线性组合的尺度和时延参数以及神经网络的权值。 相似文献
17.
神经网络经济预测法研究 总被引：10，自引：0，他引：10
本文在多层BP神经网络的基础上，结合经济类时间序列的特点，采用特殊的处理方法，建立通用的经济预测神经网络模型，并利用此模型对安徽省某一经济数据加以预测。特殊的处理方法包括前置处理、单维时间序列扩展输入节点设计、训练区数据与试验区数据划分、误差自相关神经元节点的引入，以及后置评价处理。实际的预测结果表明了该方法的先进性和可行性。 相似文献
18.
创业板公司成立时间较短、企业规模较小,其研发投入的影响因素较多,使用营业净利率、每股收益、董监高年薪、可持续增长率、资产负债率、现金流量净额和GDP这些指标,运用径向基神经网络(RBF)和逆传播神经网络(BP)方法构建一个训练完成的神经网络模型,研究发现RBF神经网络模型比BP神经网络模型具有更好的拟合、预测效果。 相似文献
19.
《Journal of The Franklin Institute》2023,360(11):7438-7456
Hybrid quantum-classical algorithms provide a promising way to harness the power of current quantum devices. In this framework, parametrized quantum circuits (PQCs) which consist of layers of parametrized unitaries can be considered as a kind of quantum neural networks. Recent works have begun to explore the potential of PQCs as general function approximators. In this work, we propose a quantum-classical deep network structure to enhance model discriminability of convolutional neural networks (CNNs). In CNNs, the convolutional layer uses linear filters to scan the input data followed by a nonlinear operation. Instead, we build PQCs, which are more potent function approximators, with more complex structures to capture the features within the receptive field. The feature maps are obtained by sliding the PQCs over the input in a similar way as CNN. We also give a training algorithm for the proposed model. Through numerical simulation, the proposed hybrid models demonstrate reasonable classification performance on MNIST and Fashion-MNIST (4-classes). In addition, we compare the performance of models in different settings. The results demonstrate that the model with high-expressibility ansaetze achieves lower cost and higher accuracy, but exhibits a “saturation” phenomenon. 相似文献
20.
The advancement in mobile technology has enabled the application of the mobile wallet or m-wallet as an innovative payment method to substitute the traditional functions of the physical wallet. However, because of pro-innovation bias, scholars have a focus on the adoption of technology and very little attention has been given to the resistance of innovation, especially in the m-wallet context. This study addressed this absence by examining the inhibitors of m-wallet innovation adoption through the lens of innovation resistance theory (IRT). By applying a sophisticated two-staged structural equation modeling-artificial neural network (SEM-ANN) approach, we successfully extended the IRT by integrating socio-demographics and perceived novelty. The study has unveiled the noncompensatory and nonlinear relationships between the predictors and m-wallet resistance. Significant predictors from SEM analysis were taken as the ANN model’s input neurons. According to the normalized importance obtained from the multilayer perceptrons of the feed-forward-back-propagation ANN algorithm, we found significant effects of education, income, usage barrier, risk barrier, value barrier, tradition barrier, and perceived novelty on m-wallet innovation resistance. The ANN model can predict m-wallet innovation resistance with an accuracy of 76.4 %. We also discussed several new and useful theoretical and practical implications for reducing m-wallet innovation resistance among consumers. 相似文献