ID 原文 译文
58378 该框架通过对被保护应用进行静态分析，提取其 native 代码的控制流特征，向开发者提供可视化策略配置视图设定关键函数，并根据策略配置生成对应的加固代码，与被保护应用的其他部分一起形成目标应用; This framework can extract the control-flow features of subroutine invocation process by static analysis，provide developers with a visual policy configuration view to set the reinforced points，generate thereinforcement code based on the CFI policy，and integrate the verification module into the target application.
58379 目标应用在运行时，通过对关键函数进行动态 CFI 检查判定是否遭遇上述攻击，从而达到保护目的. Then a CFI check is enforced during the run-time of the application to defend against the maliciousattack.
58380 实验结果表明，DroidCFI 能够通过极小的性能开销实现对应用软件 native 代码的安全性保护. Experiments show that DroidCFI can provide secure protection to native code of applications byminimal performance overhead.
58381 为解决反馈型两级交换结构( FTSA) 对调度算法的时间限制问题，提出了一种脉动反馈型两级交换结构( PFTSA) . To solve the time constraint of scheduling algorithm in the feedback-based two-stage switch architecture ( FTSA) ，a new scheme called pulsating feedback-based two-stage switch architecture ( PFTSA) is proposed，which transmits the required information back to the input port in a way of pulsating.
58382 PFTSA 将调度算法所需信息以脉动的形式反馈至输入端口，通过预处理机制使调度算法获得目标缓存的准确信息，从而避免信元冲突和信元失序. The accurate data of target buffers can be obtained by the scheduling algorithm with a preprocessingscheme，so as to avoiding the cell conflicting and disordering.
58383 相对于现有方案，PFTSA 简化了交换结构和交换流程，同时提高了时延性能 As compared to the existing schemes，PFTSA can not only simplify the switch architecture and procedure，but also improve the delay performance.
58384 针对最小二乘孪生支持向量机( LSTWSVM) 精度较低和可能存在的“奇异性”问题，提出了一种最小二乘大间隔孪生支持向量机( LSLMTSVM) . In order to overcome low accuracy and possible singularity of least squares twin support vectormachine ( LSTWSVM) ，a least squares large margin twin support vector machine ( LSLMTSVM) is presented.
58385 该算法在最小二乘孪生支持向量机的优化目标函数中引入了间隔分布，提高了算法的泛化性能. The proposed algorithm improves generalization performance by introducing margin distribution tothe optimization objective function of the LSTWSVM.
58386 在目标函数中加入正则项，实现了结构风险最小化，进一步提高了分类能力. Additionally，the structural risk minimization principle is implemented by adding the regularization term to the objective function which improves classification ability.
58387 实验结果表明，最小二乘大间隔孪生支持向量机比已有的相关算法性能更优. Experimental results show that LSLMTSVM has better classification performance than theexisting algorithm.