NETEM阅读部分的真实性分析

从真实性的角度出发,利用Bachman的任务特征框架分析了NETEM阅读理解部分对考生交际语言能力的测试.分析表明,NETEM在输入材料方面有较高的真实性,任务涉及的语言能力的真实性也有很大的提高.但从输入与回答之间的关系看,个别题目未能体现交际语言能力的特点,仍有待改进.     

基于迁移学习的卷积神经网络通道剪枝

卷积神经网络在计算机视觉等多个领域应用广泛,然而其模型参数量众多、计算开销庞大,导致许多边缘设备无法满足其存储与计算资源要求。针对其边缘部署困难,提出使用迁移学习策略改进基于BN层缩放因子通道剪枝方法的稀疏化过程。本文对比不同层级迁移方案对稀疏化效果与通道剪枝选取容限的影响;并基于网络结构搜索观点设计实验,探究其精度保持极限与迭代结构的收敛性。实验结果表明,对比原模型,采用迁移学习的通道剪枝算法,在精度损失不超过0.10的前提下,参数量减少89.1%,模型存储大小压缩89.3%;对比原剪枝方法,将剪枝阈值从0.85提升到0.97,进一步减少参数42.6%。实验证明,引入迁移策略更易实现充分的稀疏化,提高通道剪枝阈值选取容限,实现更高压缩率;并在迭代剪枝的网络结构搜索过程中,提供更高效的搜索起点,利于快速迭代趋近至搜索空间的一个网络结构局部最优解。     

LogGAN: a Log-level Generative Adversarial Network for Anomaly Detection using Permutation Event Modeling

Information Systems Frontiers - System logs that trace system states and record valuable events comprise a significant component of any computer system in our daily life. Each log contains...     

A Performance Evaluation of Classic Convolutional Neural Networks for 2D and 3D Palmprint and Palm Vein Recognition

Palmprint recognition and palm vein recognition are two emerging biometrics technologies. In the past two decades, many traditional methods have been proposed for palmprint recognition and palm vein recognition, and have achieved impressive results. However, the research on deep learning-based palmprint recognition and palm vein recognition is still very preliminary. In this paper, in order to investigate the problem of deep learning based 2D and 3D palmprint recognition and palm vein recognition in-depth, we conduct performance evaluation of seventeen representative and classic convolutional neural networks (CNNs) on one 3D palmprint database, five 2D palmprint databases and two palm vein databases. A lot of experiments have been carried out in the conditions of different network structures, different learning rates, and different numbers of network layers. We have also conducted experiments on both separate data mode and mixed data mode. Experimental results show that these classic CNNs can achieve promising recognition results, and the recognition performance of recently proposed CNNs is better. Particularly, among classic CNNs, one of the recently proposed classic CNNs, i.e., EfficientNet achieves the best recognition accuracy. However, the recognition performance of classic CNNs is still slightly worse than that of some traditional recognition methods.

     

Research on Action Strategies and Simulations of DRL and MCTS-based Intelligent Round Game

International Journal of Control, Automation and Systems - The reinforcement learning problem of complex action control in multiplayer online battlefield games has brought considerable interest in...     

Optimization of uncertain acoustic metamaterial with Helmholtz resonators based on interval model

Uncertainties existing in the acoustic metamaterial may strongly affect its unusual properties. Aiming at this actuality, the interval model is introduced to treat with uncertainties existing in the acoustic metamaterial with Helmholtz resonators. Frequency intervals in which the sound intensity transmission coefficients are certainly less than the required value and the effective bulk moduli are certainly negative are defined as conservative approximations. Frequency intervals in which the sound intensity transmission coefficients may be less than the required value and the effective bulk moduli may be negative are defined as unsafe approximations. The proportion of the conservative approximation and the unsafe approximation is defined as an approximate precision. Based on the quantification of uncertainties of the sound intensity transmission coefficients and the negative effective bulk moduli, an optimization model for the interval acoustic metamaterial with Helmholtz resonators is constructed. Numerical results showed that even suffering from effects of interval parameters, unusual properties of the optimized acoustic metamaterial (such as the bandgap of the sound transmission and the negative effective bulk modulus) could be improved.     

Attack Technology Research for Web Service

Web service is characterized by its isomerism and loose coupling;thus greatly fa- cilitate the application of operation in distributed situation, but its disadvantages cause seri- ous safety problems. The principium and characteristic of Web service classical attack tech- nology are expatiated, classical attack techniques are classified, and attack technology is sim- ulated, and then, the directions and the future of Web security are discussed.     

《中德四季晨昏杂咏》（之八）的译文鉴赏

在诗歌翻译中,译文会因为译者的不同在风格上表现出差异.然而,译文的质量不外乎经受形美、音美和意美的考量.译者应从本国读者的文化背景上结合时代的脉搏来进行“再创造”,做到形式、音律和意蕴的完美统一.歌德的诗歌《中德四季晨昏杂咏》（之八）的四种中文译本各有特色,均在不同程度上实现了“三美”的融合.     

激光熔覆含SiC金属陶瓷涂层显微组织特征

采用连续ＣＯ２激光器在钢表面进行熔覆含ＳｉＣ金属陶瓷涂层。对不同工艺条件下激光熔属层组织结构、成分与硬度进行了研究，对试验结果和凝固过程进行了传热分析。     

大截面板簧和稳定杆的热处理改进

合理选择淬火冷却介质和确定零件的材料产品质量取于受控状态和确保产品质量的提高。