【标题】Explainable Artificial Intelligence: Requirements for Explainability
【参考中译】可解释人工智能：对可解释性的要求
【类型】 期刊
【关键词】 Explainable artificial intelligence; Information; Information visualization; Explainability
【参考中译】 可解释人工智能；信息；信息可视化；可解释性
【作者】 Gayane Grigoryan
【摘要】 To date, many reasons have been suggested for making explainable artificial intelligence (XAI) models. However, it is unclear when the XAI suggested content is considered an explanation. This paper conducts a survey to determine the requirements for the information to be considered an explanation. Four minimum requirements have been prioritized based on the impact of the change they present to distinguishing between information and explainability.
【参考中译】 到目前为止，人们提出了许多理由来制作可解释的人工智能(XAI)模型。然而，尚不清楚Xai建议的内容何时被认为是一种解释。本文进行了一项调查，以确定将这些信息视为解释的要求。根据变更对区分信息和可解释性的影响，确定了四个最低要求的优先顺序。
【来源】 Proceedings of the ACM SIGSIM Conference on Principles of Advanced Discrete Simulation 2022, vol.2022
【入库时间】 2023/5/30
【标题】Development of artificial intelligence based model for the prediction of Young's modulus of polymer/carbon-nanotubes composites
【参考中译】基于人工智能的聚合物/碳纳米管复合材料杨氏模数预测模型的建立
【关键词】 AI; ML; Nanocomposites; Neural network; Polymer; Composite properties
【参考中译】 人工智能；ML；纳米复合材料；神经网络；聚合物；复合材料性能
【作者】 Nang Xuan Ho; Tien-Thinh Le; Minh Vuong Le
【摘要】 In this paper, an Artificial Intelligence (AI) model is constructed for the behavior prediction, i.e. Young's modulus, of polymer/carbon-nanotube (CNTs) composites. The AI is proposed to overcome the difficulties when studying the properties of novel composite materials, for example the time-consuming of experimental studies of resource-consuming of other numerical methods. Artificial Neural Network (ANN) model was chosen and optimized in architecture based on a parametric study. The main objective of this study is to firstly confirm that the proposed AI method performs well for nanocomposites and it can then be optimized in terms of computational time and resources in further studies. The obtained results have shown that the proposed model exhibits great performance in both training and testing phases, where the correlation coefficient is 0.986 for training part and 0.978 for the testing part.
【参考中译】 建立了聚合物/碳纳米管复合材料行为预测的人工智能模型，即杨氏S模数。人工智能的提出是为了克服研究新型复合材料性能的困难，例如其他数值方法的实验研究耗费资源的耗时。在参数研究的基础上，选择了人工神经网络(ANN)模型，并对其进行了优化。这项研究的主要目的是首先确认所提出的人工智能方法对纳米复合材料具有很好的性能，然后在进一步的研究中可以在计算时间和资源方面进行优化。结果表明，该模型在训练阶段和测试阶段都表现出了良好的性能，训练阶段的相关系数为0.986，测试阶段的相关系数为0.978。
【来源】 Mechanics of Advanced Materials and Structures 2022, vol.29, no.27
【标题】PPSS: A privacy-preserving secure framework using blockchain-enabled federated deep learning for Industrial IoTs
【参考中译】PPS：基于区块链的工业物联网联合深度学习隐私保护安全框架
【关键词】 Federated learning; Intrusion detection; Privacy preserving; Proof of learning; Security; Industrial internet of things (IIoT)
【参考中译】 联合学习；入侵检测；隐私保护；学习证明；安全；工业物联网(IIoT)
【作者】 Djallel Hamouda; Mohamed Amine Ferrag; Nadjette Benhamida; Hamid Seridi
【摘要】 The growing reliance of industry 4.0/5.0 on emergent technologies has dramatically increased the scope of cyber threats and data privacy issues. Recently, federated learning (FL) based intrusion detection systems (IDS) promote the detection of large-scale cyber-attacks in resource-constrained and heterogeneous industrial systems without exposing data to privacy issues. However, the inherent characteristics of the latter have led to problems such as a trusted validation and consensus of the federation, unreliability, and privacy protection of model upload. To address these challenges, this paper proposes a novel privacy-preserving secure framework, named PPSS, based on the use of blockchain-enabled FL with improved privacy, verifiability, and transparency. The PPSS framework adopts the permissioned-blockchain system to secure multiparty computation as well as to incentivize cross-silo FL based on a lightweight and energy-efficient consensus protocol named Proof-of-Federated Deep-Learning (PoFDL). Specifically, we design two federated stages for global model aggregation. The first stage uses differentially private training of Stochastic Gradient Descent (DP-SGD) to enforce privacy protection of client updates, while the second stage uses PoFDL protocol to prove and add new model-containing blocks to the blockchain. We study the performance of the proposed PPSS framework using a new cyber security dataset (Edge-IIoT dataset) in terms of detection rate, precision, accuracy, computation, and energy cost. The results demonstrate that the PPSS framework system can detect industrial IIoT attacks with high classification performance under two distribution modes, namely, non-independent and identically distributed (Non-IID) and independent and identically distributed (IID).
【参考中译】 工业4.0/5.0对新兴技术的日益依赖大大增加了网络威胁和数据隐私问题的范围。近年来，基于联邦学习(FL)的入侵检测系统在不暴露数据隐私问题的情况下，促进了对资源受限、异构工业系统中大规模网络攻击的检测。然而，后者的固有特性导致了联邦的可信验证和共识、模型上传的不可靠性和隐私保护等问题。为了应对这些挑战，本文提出了一种新的隐私保护安全框架PPS，该框架基于区块链使能FL的使用，提高了隐私、可验证性和透明度。PPSS框架采用允许的区块链系统来保护多方计算，并基于一种名为联合深度学习证明(PoFDL)的轻量级、节能的共识协议来激励跨竖井FL。具体地说，我们设计了两个联邦阶段来进行全局模型聚合。第一阶段使用随机梯度下降的差分私有训练(DP-SGD)来加强客户端更新的隐私保护；第二阶段使用PoFDL协议来证明并向区块链添加新的包含模型的块。我们使用一个新的网络安全数据集(Edge-IIoT数据集)从检测率、精确度、准确度、计算量和能量开销等方面研究了所提出的PPSS框架的性能。实验结果表明，在非独立同分布(Non-IID)和独立同分布(IID)两种分布模式下，PPSS框架系统能够检测到高分类性能的工业IIoT攻击。
【来源】 Pervasive and Mobile Computing 2023, vol.88