课程文献中心 更多>>

关键词： Normal distribution  

摘要： In order to capture the hidden structure of networks and reduce the dependence of community detection models on the choice of initial parameters, this paper proposes a community detection model based on hierarchical graph clustering networks with probability distribution (HGCPD). Firstly, graph convolutional networks are utilized to learn and cache feature representations of nodes in the graph. Then, a hierarchical clustering method based on node pair similarity probability is introduced to recursively construct community structures at different levels. Finally, the problem of model hyperparameter optimization is explored, and the Bayesian optimization methods are designed to automatically adjust parameters, thereby improving the efficiency of the model. Experiments on multiple network datasets of different scales show that the HGCPD model is superior to mainstream methods in terms of the accuracy and effectiveness of community detection, and the model's interpretability is verified through visualization. © 2025 Northeast University. All rights reserved.

关键词： Mobile robots  

摘要： To address the challenges of finding optimal paths in continuous spaces with graph-based planning algorithms and the low efficiency of path generation in sampling-based algorithms, this study proposes an optimal path planning algorithm based on informed sampling of convex dissection(CDI-RRT*). First, the algorithm performs convex dissection on a static map and establishes a topological graph. Guided by this graph, it uses the A* algorithm to generate an initial path, which is then optimized with an elastic band algorithm to obtain a locally optimal path. Subsequently, the CDI-RRT* constructs an initial tree within the topological framework and incorporates constraints from both convex edges and the informed set of the Informed-RRT* algorithm to build a dynamic sampling domain. By performing random sampling within this domain, the initial tree is iteratively optimized to yield an optimal path. Finally, a comparative study of the CDI-RRT* and current state-of-the-art optimal path planning algorithms is conducted through simulations and real-world experiments. Results demonstrate that the CDI-RRT* outperforms in terms of initial and optimal path generation efficiency in most scenarios, validating its feasibility and effectiveness. © 2025 Northeast University. All rights reserved.

关键词： 网络心理育人   大学生   思想政治教育   影响研究   对策  

摘要： 本研究探讨网络心理育人对大学生思想政治教育的影响.研究发现,网络心理育人提升了思政教育的吸引力和针对性,增强了大学生的自主学习能力和批判性思维,拓展了思政教育的时空维度.然而,在信息筛选、价值引导、道德风险防范、教师角色转变和评估体系建设等方面仍面临挑战.针对这些问题,本研究提出构建网络思政教育平台、创新教育方法、加强教师队伍建设、完善评估体系等对策建议,以期充分发挥网络心理育人在大学生思政教育中的积极作用.

关键词： Stiffness matrix  

摘要： The sensitivity analysis of structural Vibration modes (eigenvalues and eigenvectors) is widely used in structural Vibration control, optimal design and damage identification. At present, the main modal sensitivity analysis algorithms are modal superposition method, Nelson' s method and their improved algorithms. When these algorithms are applied to modal sensitivity analysis of large-scale engineering structures, there is generally a defect of low computational efficiency. In order to save the calculation cost, an improved subspace iteration method for modal sensitivity analysis was proposed based on approximate flexibility. Firstly, the calculation problem of modal sensitivity was transformed into the calculation problem of modal eigen-pairs after minor modification of the structure by difference Operation. Then, an approximate flexibility calculation formula was proposed to quickly estimate the inverse of the modified stiffness matrix of the structure, which was applied to the subspace iteration process to quickly obtain the modal eigen-pairs after minor modification, and accordingly the corresponding modal sensitivity could be quickly calculated. Two structural models were taken as examples to verify the proposed method. The results show that the calculation accuracy of the proposed method is basically consistent with that of the existing modal sensitivity algorithm, but the calculation time is greatly reduced. The proposed method is more suitable for the Vibration modal sensitivity analysis of large-scale engineering structures than the existing methods. © 2025 Chinese Vibration Engineering Society. All rights reserved.

关键词： 出路   互联网技术   便捷化操作   困境   消费者  

摘要： 在互联网技术快速发展的背景下,依托网络平台和智能终端设备,消费者得以实现线上订餐的便捷化操作,显著提升了消费者日常生活的便利性,并且能够为消费者提供更加丰富多样的餐食选择。在当前生活节奏较快的背景下,大量青壮年在上班、居家休息期间,更倾向于把时间花在自己感兴趣的活动上,这一消费行为变迁为网络外卖食品行业创造了新的发展契机。为保证自身能够抓住机遇,网络外卖食品需要得到更加健全的监管体系,确保行业在发展过程中能够凭借有效监管保证网络外卖食品的安全性,为消费者提供可靠保障。

关键词： 知识图谱   网络信息   自监督   强化学习   推荐模型   交互信息   特征提取   相似度计算  

摘要： 为深入理解和挖掘用户历史网络交互信息中的行为特征,动态提取用户交互行为变化,实现网络信息个性化推荐,构建了一种基于知识图谱的网络信息自监督强化学习推荐模型。该模型构建了用户网络信息交互行为的知识图谱,清晰地展示用户历史网络信息交互行为。通过基于自监督强化学习的特征提取模型,有效捕捉知识图谱中用户行为的动态变化,避免流行度偏差的负面影响,从而提取出历史网络交互信息的特征。基于知识图谱相似度计算,推荐与用户历史交互信息特征相似的网络信息实体,实现精准且个性化的推荐。实验结果验证,所提模型为用户推荐网络电影信息资源后,点击播放转化率达96.83%,网络信息个性化推荐效果明显提升。

关键词： 天府旅游名村   传播热度   TOPSIS法   熵权法   皮尔逊相关系数   乡村旅游  

摘要： 网络传播热度是衡量旅游目的地营销成效的关键指标,对于天府旅游名村而言,其创建旨在展现四川特色文旅精品,深度挖掘乡村特色文旅资源,而网络传播热度能够直观反映这些村落的旅游发展态势与游客认同程度。本研究借助数据采集软件,收集了2021—2024年首批天府旅游名村在抖音与小红书平台上发布的视频和文章,运用熵权法与TOPSIS分析法,从发文量、点赞量、收藏量、评论量四个维度对30个村落的传播热度展开评价研究,并通过皮尔逊相关性分析探究各评价指标间的关联。研究结果表明,首批天府旅游名村的网络传播热度整体处于较低水平,且在抖音平台的传播热度普遍高于小红书平台,其中通江县王坪村在抖音上、稻城亚丁村在小红书上的传播热度位居榜首,同时发文量、点赞量、收藏量和评论量之间呈现显著的正相关关系。据此,本提出针对性策略建议,仅供参考。

关键词： 海量数据场景   生成对抗网络   长短期记忆网络   推荐算法   动态序列   个性化推荐   目标函数  

摘要： 海量数据中往往包含复杂的用户行为模式、物品属性以及它们之间的关系,这些关系通常具有非线性的特点。传统的生成对抗网络(GAN)在处理序列数据时可能面临非线性建模的挑战。为有效捕捉用户的长短期兴趣变化,丰富内容多样性,并提升在海量数据场景下的处理能力和稳定性,提出一种面向海量数据场景的生成对抗网络推荐算法。长短期记忆(LSTM)网络以用户对数据场景的行为模式作为输入,输出用户感兴趣的数据场景的长短期动态序列。文中将LSTM与GAN相结合,形成L-GAN推荐模型。在该模型中,LSTM输出的长短期动态序列被输入到生成对抗网络的生成器中,通过优化损失函数生成类似真实数据场景的假样本。将假样本与真实数据场景一同输入到判别器中,通过其目标函数甄别真伪。生成器与判别器经过反复较量与训练,形成精准的推荐网络,最终输出符合用户兴趣的数据场景推荐列表。实验结果表明,所提算法在处理海量数据场景时能够准确捕捉用户的需求,进行高效且全面的个性化推荐。