Research and Practice of Hybrid Teaching for Software Testing based on Interactive Learning

To address the problems of insufficient number of personalized exercises and cases and teachers' lack of grasp of students' weak knowledge points in the current software testing online courses, we study the strategy of establishing and updating intelligent exercise sets and case libraries and analyze the answers and dig out the weak points of knowledge through group intelligence reasoning and interactive machine learning methods. This will help teachers to make uniform and targeted explanations, reduce manual judgment, and achieve intelligent teaching quality reform, and implement the educational concepts of "keeping up with the times" and "teaching according to students' abilities".     

融合共现推理的Faster R-CNN输电线路金具检测

为促进目标检测模型与电力领域业务知识有机融合,缓解金具样本间长尾分布问题,有效提升输电线路金具的自动化检测效果,提出了融合共现推理的Faster R-CNN(faster region-based convolutional neural network)输电线路金具检测模型。首先,深入研究输电线路金具结构化组合规则,通过数据驱动的方式以条件概率对金具目标间的共现连接关系进行有效表达;然后,结合图学习方法,利用学习并映射的共现概率关联作为共现图邻接矩阵,使用基线模型(Faster R-CNN)提取的特征向量作为图推理输入特征,以自学习的变换矩阵作为共现图传播权重,完成图信息传播并实现有效的共现推理模型嵌入。实验证明,融合共现推理模块的Faster R-CNN模型较原始模型提升了6.56%的准确率,对于长尾分布样本中数量较少的金具性能提升尤其显著。     

Probabilistic and interval reliability design of the composite pressure vessel

Considering the randomness or interval character of physical parameters and applied loads of composite pressure vessels (COPV), the COPV reliability-analyzing model is built. And then the computational expressions for the mean value, standard deviation and deviation of fiber stress are deduced by the random factor method. The probabilistic and interval reliability designs on COPV are implemented by utilizing the probabilistic method and interval method combined with the improved particle swarm optimization (PSO) algorithm, respectively. The influence of fluctuation of structure uncertain parameters on COPV design thickness is inspected. Finally, examples are given to demonstrate that both probabilistic and interval reliability methods can satisfy the safety requirement and both are of higher rationality than the traditional safety factor method. And probabilistic method has a lower relative error but a higher computational complexity contrasted with the interval method.     

流计算模式下概率粗糙集三支决策的快速计算

概率粗糙集三支决策是不确定问题求解的一种重要理论,流计算模式是一种新型的动态内存计算形式,实施流计算模式下三支决策的快速动态计算是一项具有挑战性的新议题。本研究以流计算模式中的两个核心计算步骤即动态增量与动态减量作为研究对象,提出了一种流计算模式下概率粗糙集三支决策域的快速动态学习方法。首先对流计算模式中三支决策动态增量和动态减量的不同变化情况进行了数据建模。然后基于不同数据变化情况分别讨论了数据增量与数据减量时三支决策域的变化推理,并且基于上述理论给出了流计算模式下的三支决策动态增减学习算法。该算法能够以更低的时间复杂度获得与经典三支决策算法相同决策效果。最后通过八种UCI数据集的实验证明了流计算模式下三支决策动态增减学习算法在时间消耗上明显优于经典概率粗糙集三支决策算法,并且在不同阈值下具有稳定的决策效率。本研究表明了流计算模式下三支决策快速计算是可行的。     

Convolutional neural networks recognition algorithm based on PCA

To improve the insufficiency of Synthetic Aperture Radar(SAR) labeled training data for Convolutional Neural Networks(CNN) and the recognition rate for large variations, a novel CNN recognition algorithm is proposed. Firstly, a set of features is extracted from the original data by unsupervised training based on PCA as the initial filter set for CNN. Secondly, in order to accelerate the training speed while avoiding over-fitting, the Rectified Linear Units(ReLU) is adopted as the non-linear function. Thirdly, to strengthen robustness and mitigate the defects of pooling upon features, a probabilistic max-pooling sampling method is introduced and local contrast normalization is exploited on features after the convolutional layer. Experiments demonstrate that our algorithm outperforms the original CNN in recognition rate and achieves better robustness for large variations and complex background.     

基于ALC的事例库结构及检索研究

为了提高基于ALC的事例检索的效率,通过概念距离及LCS推理给出了索引节点密度大且带权重的事例库层次结构,实现了事例更细致的分类及统一概念距离的语义。根据该结构的组织方式、LCS概念及概念距离给出了事例筛选及相似性度量算法,证明了通过事例筛选得到的候选事例集明显减小,采用相似性度量算法能得到合理的候选事例序列,从而提高了基于ALC的事例检索的效率。     

基于聚类和案例推理的个性化教学序列生成

针对智能教学系统中个性化教学序列生成规则难于获取的问题,提出了将案例推理技术应用于教学序列生成,设计了学生案例相似度的计算方法.为了在案例数量增大时获得较好的检索速度和命中率,提出通过关键属性聚类来对案例进行划分并设计了用于聚类的惰性2-中心点算法,最后以实例验证了系统设计的可用性.     

一种改进的群目标自适应跟踪算法

为提高对群目标在机动情况下的跟踪性能,提出一种改进的群目标自适应跟踪算法．在群质心状态估计中,在修正“当前”统计模型的基础上,利用群质心的速度预测和速度估计的偏差进行过程噪声方差自适应调整,并引入强跟踪滤波中的渐消因子,实时调节群质心的状态预测协方差．在扩展状态估计中,将其对应的椭圆面积预测值和估计值的偏差以及偏差变化率作为模糊输入量,采用模糊推理法自适应输出扩展状态的预测参数．此外,提供了群目标分裂机动的判决方法．仿真结果表明,与现有方法相比,本文算法增强了对群目标在突发机动时的自适应跟踪能力,并能有效检测出群的分裂机动．     

计及输入变量强相关性的概率潮流计算模型

利用统计学的方法,以风电场出力为例,对其大量历史数据进行相关性分析,将具有强相关性的随机变量用最小二乘进行线性拟合,并用其中一个变量的线性表达式表示,从而消去随机变量之间的强相关性,并认为余下的呈弱相关或不相关的随机变量之间相互独立,以便采用半不变量法计算概率潮流。