ID 原文 译文
17905 为了解决基于集中式算法的传统物联网数据分析处理方式易引发网络带宽压力过大、延迟过高以及数据隐私安全等问题，该文针对弹性网络回归这一典型的线性回归模型，提出一种面向物联网(IoT)的分布式学习算法。 In order to solve the problems caused by the traditional data analysis based on the centralized algorithm in the IoT, such as excessive bandwidth occupation, high communication latency and data privacy leakage, considering the typical linear regression model of elastic net regression, a distributed learning algorithm for Internet of Things (IoT) is proposed in this paper.
17906 该算法基于交替方向乘子法(ADMM)，将弹性网络回归目标优化问题分解为多个能够由物联网节点利用本地数据进行独立求解的子问题。 This algorithm is based on the the Alternating DirectionMethod of Multipliers (ADMM) framework. It decomposes the objective problem of elastic net regression intoseveral sub-problems that can be solved independently by each IoT node using its local data.
17907 不同于传统的集中式算法，该算法并不要求物联网节点将隐私数据上传至服务器进行训练，而仅仅传递本地训练的中间参数，再由服务器进行简单整合，以这样的协作方式经过多轮迭代获得最终结果。 Different from traditional centralized algorithms, the proposed algorithm does not require the IoT node to upload its private data to the server for training, but rather the locally trained intermediate parameters to the server for aggregation. In such a collaborative manner, the server can finally obtain the objective model after severaliterations.
17908 基于两个典型数据集的实验结果表明：该算法能够在几十轮迭代内快速收敛到最优解。 The experimental results on two typical datasets indicate that the proposed algorithm can quicklyconverge to the optimal solution within dozens of iterations.
17909 相比于由单个节点独立训练模型的本地化算法，该算法提高了模型结果的有效性和准确性； As compared to the localized algorithm in whicheach node trains the model solely based on its own local data, the proposed algorithm improves the validity andthe accuracy of training models;
17910 相比于集中式算法，该算法在确保计算准确性和可扩展性的同时，可有效地保护个体隐私数据的安全性。 as compared to the centralized algorithm, the proposed algorithm canguarantee the accuracy and the scalability of model training, and well protect the individual private data fromleakage.
17911 软件定义光网络(SDON)作为智能光网络中最新一代网络架构，其控制平面承载着诸多核心功能，其中控制平面的生存性、控制冗余和控制时延等因素对网络整体性能起到至关重要的作用。 Software-Defined Optical Network (SDON) is the latest generation network architecture in intelligentoptical networks. Its control plane carries many core functions. The survivability of control plane, control redundancy and control delay are crucial to the overall performance of the network.
17912 该文提出一种以生存性条件为约束的软件定义光网络(SCD)控制器部署算法，在保证用户对网络生存性需求的前提下，利用最短路径和极小支配集等数学原理来降低控制时延和减少控制器部署个数，降低控制冗余，并利用联合判决条件选择管控中心部署节点，协调控制器间的工作。 In this paper, aSurvivability-Constrained software-Defined (SCD) optical network controller deployment algorithm is proposed.Under the premise of ensuring users' network survivability requirements, mathematical principles such asshortest path and minimum dominance set are used to reduce control delay, and reduce the number ofcontroller deployments to reduce control redundancy. A joint judgment condition is used to select the controlcenter deployment node to coordinate the work between the controllers.
17913 实验表明：首先，所提算法可以百分之百保证用户对网络的生存性要求； Experiments show that: Firstly, theproposed algorithm can guarantee the user's survivability requirements for the network 100%.
17914 其次，所提算法相对于C-MPC算法至少降低了15%的网络故障告警概率，提高了网络生存性； Secondly, theproposed algorithm reduces the network failure alarm probability by at least 15% compared with the C-MPCalgorithm, and improves the network survivability.