基于强化学习的渗透路径推荐模型

渗透测试的核心问题是渗透测试路径的规划,手动规划依赖测试人员的经验,而自动生成渗透路径主要基于网络安全的先验知识和特定的漏洞或网络场景,所需成本高且缺乏灵活性。针对这些问题,提出一种基于强化学习的渗透路径推荐模型QLPT,通过多回合的漏洞选择和奖励反馈,最终给出针对渗透对象的最佳渗透路径。在开源靶场的渗透实验结果表明,与手动测试的渗透路径相比,所提模型推荐的路径具有较高一致性,验证了该模型的可行性与准确性;与自动化渗透测试框架Metasploit相比,该模型在适应所有渗透场景方面也更具灵活性。     

基于多媒体技术的认知功能评估系统之研究

基于多媒体技术,结合MSP430,设计了一款交互式的认知功能评估系统,该系统主要由MSP430微控制器、电磁感应模块、3轴传感器、LCD和无线通信模块组成。开发了包括命名、记忆、注意、抽象、定向、执行的6类游戏来评估大脑认知功能。实验中使用科技接受模型对系统进行评估（满分为7分,从“非常不同意”到“非常同意”分数依次增加）,其中使用意愿得分6.52分、娱乐性得分6.45分。实验结果表明：正常老人和轻度认知障碍老人相比,游戏执行时间更短（平均相差27.26 s）、准确率更高（平均相差4.38%）、数字记忆广度更好（平均相差0.7分）。     

基于CSCL的多媒体学习系统

本文分析了国内外计算机支持的协同学习(Computer Supported Cooperative Learning,CSCL)系统开发应用的现状,指出现有CSCL系统缺陷,并提出了基于Web的多媒体学习系统的构架、实现原理与关键技术。     

优化强化学习路径特征分类的脉象识别法

脉象识别是中医诊断的重要手段之一。长期以来,依据个人经验进行的脉诊制约了中医的推广与发展。因此,利用传感设备进行脉象识别的研究正在逐步展开。针对神经网络识别脉象的相关研究中,存在需要大量训练数据集,以及存在处理“黑箱”和时间花销较大等问题,在强化学习的框架下,提出了一种采用马尔可夫决策和蒙特卡罗搜索的脉象图分析法。首先依据中医理论对特定的脉象进行路径分类,然后在此基础上为不同的路径选择代表性特征,最终通过对代表性特征的阈值对比完成对脉象的识别。实验结果表明,所提方法可缩减训练时间和所需资源,并可保留完整的经验轨迹;且在提高脉象识别的准确率的同时,还可解决数据处理过程中的“黑箱”问题。     

基于故障树的等级测评专家系统模型研究*

结合等级测评、风险评估和专家系统的相关技术,提出了一种将故障树和专家系统技术用于等级测评结果的综合分析模型。该模型通过将评价标准转换成规则固化在知识库中,利用故障树对知识进行展示和分析,并利用不确定推理方法对评估事实进行推理,得到该信息系统存在的安全问题可能会发生什么安全事件,从而紧密地将等级保护要求和用户安全需求联系在一起,极大地提高了用户实践等级保护的积极性,并为等级测评与风险评估有机结合提供了一种新的思路。另外,本系统还利用粗糙集理论从信息安全事件中自动获取推理规则,使得推理规则能随着安全状态的变化     

How flipped learning based on the cognitive theory of multimedia learning affects students' academic achievements

The study aims to identify the effect of a flipped classroom approach designed according to the cognitive theory of multimedia learning on the academic achievements of eighth‐grade students (aged 14 years) in Saudi Arabia in computer science. To this end, a quasi‐experimental design was used, with a sample of 67 students; 33 students were assigned to the experimental group, whereas 34 comprised the control group. The experimental group was subjected to the flipped classroom approach, whereas the control group was given direct instruction. To measure student achievements, an instrument that measures cognitive skills based on Revised Bloom's taxonomy levels was designed. Findings revealed a positive effect on the experimental group's achievement levels with respect to Bloom's higher order thinking skills, that is, applying, analysing, and evaluating. No difference was found between the two groups in terms of academic achievements at the remembering and understanding levels of Bloom's taxonomy. Moreover, learners with low prior knowledge showed a higher improvement in academic achievements compared with those with high prior knowledge. This corresponds to the assumption of the cognitive theory of multimedia learning that learners with low prior knowledge would benefit from its principles more than learners with high prior knowledge.     

一种用于多媒体学习系统的权限控制算法

随着计算机网络在教育教学中的广泛应用,CSCL[5](Computer Supported Collaborative Learning)越来越显示它的重要性,同时权限控制也倍受专家们的关注.本文主要讨论基于CCSL的多媒体学习系统中权限控制,提出了多媒体学习系统中权限控制的具体实施策略与实现算法.     

基于试题关系树的自适应诊断测试模型

针对实际测试过程中试题可能包含多个知识点组合的问题,提出一种基于试题关系树的自适应诊断测试模型的解决方案.根据知识点组合的包含关系,分析得出试题之间存在的3种前提关系,在此基础上提出以与子结点和或子结点将试题库进行关联,形成试题关系树.重点分析了基于试题库构建试题关系树的过程,给出了基于试题关系树的测试策略与诊断过程,并最终形成诊断决策树.实验结果表明,该模型可有效地减小平均测试长度,提高诊断效率.     

Efficient multimedia information mining framework based on deep learning and self-organizing model

Multimedia Tools and Applications - With the progress of society and the acceleration of urbanization, the problem of multimedia assisted urban traffic is becoming increasingly apparent....     

Development path of basic education based on 5G technology and multimedia embedded system

At the moment, the reason is given great interest, pioneering in various fields and point exposure to 5G technology in terms of e-learning. Online learning experience is about existing technology and online bachelor's degree. The survey is about what students are offered that can be carried out with the development of 5G technology in the future just provided. For this purpose, the main multimedia application properties are considered. An analysis of the literature and numerical evaluation is obtained, compared to LTE, which is based on the performance services provided in theory and 5G analysis and simulation results. Education K-12, higher education, trying to find the best way to educate students. The increased amount of data is accepted by 5G (or delayed) due to the lower latency, which can be an excellent capability for VR or other hyper-realistic support experience. Support for embedded systems, their language and culture extends to the platform provides an interface of building blocks needed by other developers.     

Forecasting air passenger traffic flow based on the two-phase learning model

The future airports will head toward a highly intelligent direction, like the unmanned check-in services, while the scale and resources allocation of the ground service are tightly related to the air passenger flow. Therefore, forecasting passenger flow accurately will affect the development of future airports and the optimization of service of civil airlines significantly. As a kind of time series, air passenger flow is influenced by multiple factors, particularly, the stochastic part of seasonality, trend and volatility. These will ultimately affect the accuracy of the prediction. Therefore, this paper introduces a prediction model based on a two-phase learning framework. In phase one, various predictors cope with different features of time series in parallel and the prediction results are integrated in phase two. Furthermore, this paper has compared principal error indicators with actual data and results show that the two-phase learning model performs better than current fusion models and owns stable performance.

     

Network intrusion detection based on deep learning model optimized with rule-based hybrid feature selection

ABSTRACT

Network Intrusion Detection System (NIDS) is often used to classify network traffic in an attempt to protect computer systems from various network attacks. A major component for building an efficient intrusion detection system is the preprocessing of network traffic and identification of essential features which is essential for building robust classifier. In this study, a NIDS based on deep learning model optimized with rule-based hybrid feature selection is proposed. The architecture is divided into three phases namely: hybrid feature selection, rule evaluation and detection. Several search methods and attribute evaluators were combined for features selection to enhance experimentation and comparison. The results obtained showed that the number of selected features will not affect the detection accuracy of the feature selection algorithms, but directly proportional to the performance of the base classifier. Results from the performance comparison proved that the proposed method outperforms other related methods with reduction of false alarm rate, high accuracy rate, reduced training and testing time of 1.2%, 98.8%, 7.17s and 3.11s, respectively. Finally, the simulation experiments on standard evaluation metrics showed that the proposed method is suitable for attack classification in NIDS.     

基于模糊综合评判的网络认知模型

针对传统传输控制协议(TCP)应用于异构网络的局限性,提出了一种基于模糊综合评判的网络认知模型。该模型通过建立隶属度函数和不同网络环境下的动态权重分布,运用模糊综合评判的方法来区分无线误码丢包和网络拥塞丢包。仿真实验证明：与传统TCP协议相比,该模型在不同的网络条件下,能够较准确地区分无线误码丢包和网络拥塞丢包,提高了TCP的吞吐量,改善了网络性能。     

基于扩展攻击树的信息系统安全风险评估

在攻击树模型的基础上,对该模型进行扩展,提出了一种新的定量的风险评估方法.在对叶子节点(原子攻击)风险值的量化中,采用了多属性效用理论,使得评估更加客观;对该方法的每一步骤均给出了具体的算法,为实现自动化的评估工具建立了基础.     

基于Petri网的知识空间建模方法与学习路径控制

对学生学习路径的控制和学习状态的了解在自主学习系统中是一个重要的问题,以知识空间理论为基础建立了学习状态空间,并通过Petri网对该学习状态空间进行了建模,利用Petri网实现了学习状态空间中所有可能的学习路径控制和学生学习状态的了解,在此模型中学生可以利用Petri网的变迁实现自由的学习控制,并通过Petri网库所标识来了解学生的学习状态,达到个性化学习路径的目的。     

基于Petri网工作流模型展开树的路径序列相似性算法

在实际的数据迁移项目中,为了解决数据映射的问题,需要确定两个工作流模型之间的相似度。从工作流模型的相似性方面进行分析阐述,提出了基于Petri网的工作流模型展开树的路径序列相似性算法,首先采用深度优先搜索算法和动态规划算法对模型进行搜索,其次通过提出的算法获取展开树的所有路径序列,最后利用编辑距离算法计算两个模型序列之间的两两相似度,进而完成模型相似性计算,相较于其他的主流相似度算法,主要优点在于可以精确计算得到模型部分结构和行为相似度,可以更好的确定流程间映射,从而找到数据映射的解决方法。实验结果表明：该方法较主流的基于模型结构和行为相似性算法,计算合理性和准确性有很大提升。     

Dynamic behavioral assessment model based on Hebb learning rule

Behavioral assessment based on computing system is with important value for computer-simulated training and system diagnosis. However, the existing assessment is a static method for ex post evaluation, and the low efficiency and high complexity have been the urgent problems to be solved in the academic field. In this paper, we propose an adaptive dynamic behavioral assessment model based on Hebb learning rule that effectively combines the assessment standard and the weights of factors. The dynamic behavioral assessment considers the relative weights between the assessment indexes, whereas the existing assessment method does not; the dynamic behavioral assessment uses the assessment standard data recursively and can conduct an instant assessment for the objectives. We have built an assessment system for computer-simulated training, and took the pilot behavioral assessment for example to test and verify the dynamic behavioral assessment mode. Experimental results show that the dynamic behavioral assessment model based on Hebb learning rule has more advantage in assessment efficiency and online computing support.

     

基于强化学习的快速探索随机树特殊环境中路径重规划算法

针对移动机器人在未知的特殊环境(如U型、狭窄且不规则通道等)下路径规划效率低问题,本文提出一种强化学习(RL)驱动快速探索随机树(RRT)的局部路径重规划方法(RL-RRT).该方法利用Sarsa(λ)优化RRT的随机树扩展过程,既保持未知环境中RRT的随机探索性,又利用Sarsa(λ)缩减无效区域的探索代价.具体来说,在满足移动机器人运动学模型约束的同时,通过设定扩展节点的回报函数、目标距离函数和平滑度目标函数,缩减无效节点,加速探索过程,从而达到路径规划多目标决策优化的目标.仿真实验中,将本方法用于多种未知的特殊环境,实验结果显示出RL-RRT算法的可行性、有效性及其性能优势.