摘要
由一名新闻记者-机器人与机器学习每日新闻编辑-调查人员讨论人工智能的新发现。根据NewsRx记者在法国雷恩的新闻报道，研究表明：“汽车交通事故是对全球公共安全的重大威胁，每年造成大量伤亡。本文介绍了一种全面、可解释的人工智能(XAI)工件设计，集成事故数据，供不同利益相关者和决策者使用。”新闻记者从雷恩商学院的研究中获得了一句话，“它提出了负责任的、解释性的和可解释性的模型，并用系统级的分类法对与不同伤害严重程度相关的驾驶相关行为进行分类，从而在理论上为可解释的分析做出了贡献。”我们采用了各种先进的技术，如随机数据缺失(MAR)和贝叶斯动态条件输入解决缺失记录，合成少数群体过抽样技术解决数据不平衡问题，分类提升(CatBoost)和沙普利相加解释(SHAP)确定和分析危险因素对伤害严重程度的重要性和依赖性。为了揭示影响交通事故和伤害严重程度的潜在时空因素，我们在第二阶段开发了几个预测模型，包括极端梯度Boosting(XGBoost)、随机森林(RF)、深度神经网络(DNN)和参数微调，并采用S-HAP方法，采用模型不可知解释技术将事故和伤害严重程度从模型中分离出来。我们对跨功能类别的系统级分类进行了分析和总结。
Abstract
By a News Reporter-Staff News Editor at Robotics & Machine Learning Daily News Daily News-Investigators discuss new findings in Artificial Intelligence. According to news reporting originating in Rennes, Fran ce, by NewsRx journalists, research stated, "Automobile traffic accidents repres ent a significant threat to global public safety, resulting in numerous injuries and fatalities annually. This paper introduces a comprehensive, explainable art ificial intelligence (XAI) artifact design, integrating accident data for utiliz ation by diverse stakeholders and decision-makers." The news reporters obtained a quote from the research from the Rennes School of Business, "It proposes responsible, explanatory, and interpretable models with a systems-level taxonomy categorizing aspects of driver-related behaviors associa ted with varying injury severity levels, thereby contributing theoretically to e xplainable analytics. In the initial phase, we employed various advanced techniq ues such as data missing at random (MAR) with Bayesian dynamic conditional imput ation for addressing missing records, synthetic minority oversampling technique for data imbalance issues, and categorical boosting (CatBoost) combined with SHa pley Additive exPlanations (SHAP) for determining and analyzing the importance a nd dependence of risk factors on injury severity. Additionally, exploratory feat ure analysis was conducted to uncover hidden spatiotemporal elements influencing traffic accidents and injury severity levels. We developed several predictive m odels in the second phase, including eXtreme Gradient Boosting (XGBoost), random forest (RF), deep neural networks (DNN), and fine-tuned parameters. Using the S HAP approach, we employed model-agnostic interpretation techniques to separate e xplanations from models. In the final phase, we provided an analysis and summary of the system-level taxonomy across feature categories."
NETL
NSTL