[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"item-2336":3,"related-2336":69},{"id":4,"title":5,"url":6,"summary":7,"summary_zh":8,"content":9,"source_name":10,"source_url":6,"published_at":11,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":15,"score_detail":16,"sources":18,"tags":20,"primary_topic":9,"secondary_topics":24,"classification":25,"search_phrases":47,"slug":50,"view_count":42,"doi":51,"paper":52,"created_at":68},2336,"System architecture and interactive LLM prompt framework for the integration of clinical and mitochondrial data in obesity: a software prototype and feasibility study","https:\u002F\u002Fdoi.org\u002F10.3389\u002Ffrai.2026.1923426","Introduction Obesity is a critical global health challenge due to its close association with cardiovascular and metabolic risk. At the molecular level, the lipotoxic environment alters the structure, dynamics, and connectivity of the mitochondrial network. Despite advances in cell biology and digital health, there is a lack of lightweight computational tools capable of integrating clinical records with molecular and photomicrographic data in a standardized way. This study presents the design, development, and feasibility assessment of a decoupled software prototype and an interactive prompt engineering framework for integrating clinical and histological mitochondrial data in molecular obesity research. Methodology A client-side software architecture was designed using HTML5, CSS3, and the IndexedDB API for the transactional and anonymized persistence of medical records (coded using the ICD-10 classification) and micrographs of human placental tissue (obtained using transmission electron microscopy, immunofluorescence, and immunohistochemistry for the evaluation of mitochondrial fusion molecular markers such as Mitofusin-2, nitrotyrosine). For interpretive processing, a framework of prompts structured under the Chain-of-Thought paradigm was built, designed to interact with Large-Scale Language Models (LLMs). The interface’s usability, local storage performance, and the consistency of AI-assisted reasoning were evaluated. Results The developed platform enabled the seamless and instantaneous management of independent patient tabs (unique storage keys mitochondria_001 to 010), achieving local read\u002Fwrite latency without reliance on external servers or transmission of sensitive personal data. The prompt framework proved effective in guiding LLMs in translating mitochondrial biomarkers and clinical data into structured summaries of oxidative stress status and metabolic risk. The lack of a trained computer vision model for automatic segmentation is declared as the main limitation of the current study. Conclusion The software prototype and prompt interface provide a low-cost, secure (data privacy-oriented), and clinically intuitive solution for translational data management in obesity. This work demonstrates the feasibility of integrating heterogeneous biomedical parameters using conversational AI tools and local storage, laying the groundwork for the future incorporation of image segmentation algorithms.","引言 肥胖因其与心血管和代谢风险的密切关联，已成为全球性的重大健康挑战。在分子层面，脂毒性环境会改变线粒体网络的结构、动力学及连接性。尽管细胞生物学和数字健康领域取得了进展，但目前仍缺乏能够以标准化方式将临床记录与分子及显微图像数据整合的轻量级计算工具。本研究介绍了一种解耦软件原型和交互式提示工程框架的设计、开发及可行性评估，旨在整合分子肥胖研究中的临床与组织学线粒体数据。方法 采用HTML5、CSS3和IndexedDB API设计了客户端软件架构，用于医学记录（采用ICD-10分类编码）和人类胎盘组织显微图像（通过透射电子显微镜、免疫荧光和免疫组织化学获得，用于评估线粒体融合分子标志物如Mitofusin-2、硝基酪氨酸）的事务性匿名持久化存储。在解释性处理方面，构建了基于思维链（Chain-of-Thought）范式结构的提示框架，旨在与大语言模型（LLMs）进行交互。对界面的可用性、本地存储性能以及AI辅助推理的一致性进行了评估。结果 所开发的平台实现了独立患者标签页（唯一存储键mitochondria_001至010）的无缝即时管理，在无需依赖外部服务器或传输敏感个人数据的情况下，实现了本地读写延迟。提示框架在引导LLMs将线粒体生物标志物和临床数据转化为氧化应激状态和代谢风险的结构化摘要方面被证明有效。缺乏用于自动分割的经过训练的计算机视觉模型被声明为本研究的主要局限性。结论 该软件原型和提示界面为肥胖领域的转化数据管理提供了一种低成本、安全（以数据隐私为导向）且临床直观的解决方案。本研究证明了使用对话式AI工具和本地存储整合异构生物医学参数的可行性，为未来纳入图像分割算法奠定了基础。",null,"Frontiers in Artificial Intelligence","2026-09-10T00:00:00Z","论文",10,false,0,{"impact":15,"substance":15,"depth":15,"authority":15,"freshness":15,"relevant":15,"comment":17},"该研究聚焦肥胖临床与线粒体数据整合的软件原型，属医学信息学领域，与三农、农业信息化、智慧农业无直接关联，不建议进入每日精选。",[19],{"name":10,"url":6},[21,22,23],"农业人工智能","数字健康","数据隐私",[],{"status":26,"locked":14,"version":27,"method":28,"evidence":29},"needs_review","2026-10-22.2","rules",{"candidates":30,"agricultural_evidence":14,"reason":46},[31,40],{"code":32,"score":33,"supported":34,"title":35,"abstract":36,"tags":39},"I2",7,true,[],[37,38],"计算机视觉","computer vision",[21],{"code":41,"score":42,"supported":14,"title":43,"abstract":44,"tags":45},"I6",1,[],[],[23],"农业相关性需要复核",[48,49],"农业人工智能 数字健康 数据隐私","农业人工智能 数字健康","农业人工智能数字健康数据隐私-2336","10.3389\u002Ffrai.2026.1923426",{"doi":51,"openalex_id":53,"authors":54,"venue":10,"cited_by_count":15,"oa_url":60,"card":61,"direction":65,"ingested_from":67},"W7212155685",[55,58],{"name":56,"orcid":57},"Karenth Milena Rodríguez-Córdoba","https:\u002F\u002Forcid.org\u002F0000-0003-4010-2813",{"name":59,"orcid":9},"William Darío Ávila-Díaz","https:\u002F\u002Fwww.frontiersin.org\u002Fjournals\u002Fartificial-intelligence\u002Farticles\u002F10.3389\u002Ffrai.2026.1923426\u002Fpdf",{"tldr":62,"method":63,"finding":64,"direction":65,"opportunity":66},"开发轻量本地软件原型与LLM提示框架，整合肥胖临床与线粒体数据。","HTML5\u002FIndexedDB本地存储临床与电镜图像，结合思维链提示调用大语言模","实现无服务器、隐私安全的异构数据管理，提示框架可有效生成氧化应激与代谢风险摘要。","农业人工智能与决策模型","可迁移至农业场景，构建本地化多模态数据管理与LLM辅助决策工具，并补足自动图像分割能力。","openalex","2026-09-13T23:30:41.079545Z",{"total":70,"page":42,"page_size":70,"items":71},6,[72,143,207,268,334,386],{"id":73,"title":74,"url":75,"summary":76,"summary_zh":77,"content":9,"source_name":78,"source_url":75,"published_at":79,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":80,"score_detail":81,"sources":88,"tags":90,"primary_topic":94,"secondary_topics":99,"classification":100,"search_phrases":123,"slug":126,"view_count":15,"doi":127,"paper":128,"created_at":142},5192,"A Comparative Study of Client-Level Stabilization Techniques in Federated Learning for Robust Plant Disease Recognition","https:\u002F\u002Fdoi.org\u002F10.3390\u002Fs26206566","The successful integration of artificial intelligence (AI) algorithms in precision agriculture provided farmers with the means to identify potential issues, such as early disease detection or pest infestations, by making crop-related predictions. The new technological developments in this field improve the functioning of traditional agriculture by reducing the dependence on specialized human expertise and facilitating greater access to diagnostic procedures. However, most studies conducted in this area rely on centralized training of AI models, which leads to increased costs and raised privacy-related concerns. Federated learning (FL) offers a compelling framework for collaboratively training machine learning (ML) models on distributed datasets without centralizing the data owned by each participant. FL offers significant advantages in the preservation of data privacy, facilitates the collaborative exploitation of geographically distributed data sources, and reduces the risks associated with centralized collection. In this paper, we investigate the current limitations and performance boundaries of existing Federated and privacy-preserving learning in distributed sensing systems for plant disease detection, in scenarios that closely resemble real-world conditions. Our analysis specifically examined how FL aggregation algorithms, such as FedAvg, FedAvgM and FedProx, work in a custom multi-client experimental setup where data is partitioned so that each participant observes only those symptom patterns associated with specific environmental and climatic conditions. We propose further partial aggregation of the local model, an approach where we send only the model backbone to the central server, while the classification head of each participant remains strictly on the client side. The construction of environmentally heterogeneity is combined with a globally representative stabilizer node designed to preserve knowledge across disease classes that may be absent from certain clients.","人工智能（AI）算法在精准农业中的成功集成，为农民提供了通过作物相关预测来识别潜在问题（如早期病害检测或虫害侵染）的手段。该领域的新技术发展通过减少对专业人类知识的依赖并促进更广泛的诊断程序获取，改善了传统农业的运作方式。然而，该领域的大多数研究依赖于AI模型的集中式训练，这导致成本增加并引发隐私相关担忧。联邦学习（FL）提供了一个引人注目的框架，可在分布式数据集上协作训练机器学习（ML）模型，而无需集中化各参与者拥有的数据。联邦学习在保护数据隐私方面具有显著优势，促进了地理上分布的数据源的协作利用，并降低了与集中收集相关的风险。本文研究了在高度接近真实条件的场景中，现有联邦学习与隐私保护学习在用于植物病害检测的分布式感知系统中的当前局限性与性能边界。我们的分析特别考察了联邦学习聚合算法（如FedAvg、FedAvgM和FedProx）在自定义多客户端实验设置中的工作方式，其中数据被划分，使得每个参与者仅观察到与特定环境和气候条件相关的症状模式。我们提出对本地模型进行进一步的部分聚合，即仅将模型骨干发送至中央服务器，而每个参与者的分类头严格保留在客户端侧。环境异质性的构建与一个全局代表性的稳定器节点相结合，该节点旨在保留某些客户端可能缺失的跨病害类别的知识。","Sensors","2026-10-16T00:00:00Z",78,{"impact":82,"substance":83,"depth":84,"authority":85,"freshness":86,"relevant":42,"comment":87},18,21,17,13,9,"该研究针对植物病害识别中的联邦学习客户端稳定化方法提出部分聚合与稳定节点方案，方法新颖、实验贴近真实场景，对智慧农业隐私计算有参考价值，值得进入每日精选。",[89],{"name":78,"url":75},[91,21,92,93,23],"智慧农业","联邦学习","植物病害识别",{"code":32,"slug":95,"name":96,"path":97,"parent_code":98},"agricultural-ai","农业AI与大模型","\u002Ftopic\u002Fdigital-agriculture\u002Fagricultural-ai","I",[],{"status":101,"locked":14,"version":27,"method":28,"evidence":102},"classified",{"candidates":103,"agricultural_evidence":34},[104,112,119],{"code":32,"score":105,"supported":34,"title":106,"abstract":107,"tags":111},12,[],[108,109,110],"机器学习","machine learning","artificial intelligence",[21,93,92],{"code":113,"score":33,"supported":34,"title":114,"abstract":115,"tags":118},"I1",[],[116,117],"精准农业","precision agriculture",[91],{"code":41,"score":42,"supported":14,"title":120,"abstract":121,"tags":122},[],[],[23],[124,125],"联邦学习 植物病害识别","FedProx 作物病害检测","联邦学习植物病害识别-5192","10.3390\u002Fs26206566",{"doi":127,"openalex_id":129,"authors":130,"venue":78,"cited_by_count":15,"oa_url":75,"card":137,"direction":65,"ingested_from":67},"W7222586734",[131,134],{"name":132,"orcid":133},"Aurora-Felicia Cristea","https:\u002F\u002Forcid.org\u002F0009-0003-9718-065X",{"name":135,"orcid":136},"Ciprian Dobre","https:\u002F\u002Forcid.org\u002F0000-0003-4638-7725",{"tldr":138,"method":139,"finding":140,"direction":65,"opportunity":141},"比较联邦学习客户端级稳定化技术，提升植物病害识别的鲁棒性。","FedAvg\u002FFedAvgM\u002FFedProx聚合，仅上传骨干网络、分类头留本地，","部分聚合加稳定节点可缓解环境异质性与类别缺失带来的性能退化。","可探索面向非独立同分布农业数据的个性化联邦学习与通信效率优化。","2026-10-17T23:30:45.845684Z",{"id":144,"title":145,"url":146,"summary":147,"summary_zh":148,"content":9,"source_name":149,"source_url":146,"published_at":150,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":15,"score_detail":151,"sources":153,"tags":155,"primary_topic":9,"secondary_topics":158,"classification":159,"search_phrases":179,"slug":182,"view_count":183,"doi":184,"paper":185,"created_at":206},3858,"Applications of artificial intelligence in personalized nutrition interventions","https:\u002F\u002Fdoi.org\u002F10.3389\u002Ffnut.2026.1974199","Unhealthy dietary patterns remain a major modifiable contributor to the global burden of obesity, type 2 diabetes mellitus (T2DM), cardiovascular disease, and cancer. However, individuals can respond markedly differently to identical foods, limiting the effectiveness of one-size-fits-all dietary guidance. This review examines the application of artificial intelligence (AI) to personalized nutrition interventions across three interconnected layers. At the biological level, inter-individual variability can be characterized by three mechanistic dimensions: nutrigenomic variation, which influences nutrient metabolism and disease susceptibility; metabolomic biomarkers, which objectively capture dietary exposure and early metabolic alterations; and the gut microbiome, which bidirectionally mediates diet–host interactions. At the technological level, AI integrates these dimensions into multimodal digital profiles. Machine learning models predict individual postprandial glycemic and lipemic responses, with reported correlations of up to r = 0.77 in large cohorts; deep learning and computer vision automate dietary assessment; natural language processing and large language models (LLMs) support nutrition counseling; knowledge graphs enable constraint-aware food recommendations; and digital-twin and adaptive closed-loop systems incorporating continuous glucose monitoring and wearable-device data support real-time nutritional interventions. However, the strength of clinical evidence varies across disease domains. Glycemic management in T2DM and prediabetes currently has the most mature randomized controlled trial evidence: an AI-enabled personalized intervention combining algorithmic dietary personalization with CGM-linked feedback improved glycemic endpoints compared with a Mediterranean diet, while a fully automated AI-delivered diabetes prevention program achieved outcomes non-inferior to those of human coaching. In obesity, cardiovascular disease, oncology, and gastrointestinal disorders, the evidence remains mixed or preliminary. Clinical translation also faces substantial challenges, including biased and inconsistently standardized dietary and multi-omics data; the underrepresentation of non-European populations, which limits model generalizability and raises concerns about health equity; limited model explainability; privacy and data-governance challenges; immature regulatory pathways for adaptive software as a medical device; and specific risks associated with LLMs, such as hallucinated or guideline-discordant dietary advice. AI-enabled personalized nutrition is therefore progressing from proof of concept toward clinical translation. Realizing its potential equitably will require multi-ethnic nutrition–omics cohorts, privacy-preserving multicenter collaboration, dietitian-led human–AI collaboration, and lifecycle safety governance for generative models.","不健康的饮食模式仍是全球肥胖、2型糖尿病(T2DM)、心血管疾病和癌症负担的主要可改变因素。然而，个体对相同食物的反应可能显著不同，这限制了一刀切式饮食指导的有效性。本综述从三个相互关联的层面考察人工智能(AI)在个性化营养干预中的应用。在生物学层面，个体间差异可通过三个机制维度加以表征：营养基因组学变异，其影响营养素代谢和疾病易感性；代谢组学生物标志物，其可客观反映膳食暴露和早期代谢改变；以及肠道微生物组，其双向介导饮食–宿主相互作用。在技术层面，AI将这些维度整合为多模态数字画像。机器学习模型可预测个体餐后血糖和血脂反应，在大型队列中报告的相关性高达r = 0.77；深度学习和计算机视觉可自动化膳食评估；自然语言处理和大语言模型(LLMs)可支持营养咨询；知识图谱可实现约束感知的食物推荐；结合连续血糖监测和可穿戴设备数据的数字孪生和自适应闭环系统可支持实时营养干预。然而，临床证据强度在不同疾病领域存在差异。T2DM和糖尿病前期的血糖管理目前拥有最成熟的随机对照试验证据：一项AI赋能的个性化干预将算法驱动的饮食个性化与CGM关联反馈相结合，与地中海饮食相比改善了血糖终点，而一项完全自动化、由AI实施的糖尿病预防项目取得了不劣于人工指导的结果。在肥胖、心血管疾病、肿瘤学和胃肠道疾病中，证据仍不一致或处于初步阶段。临床转化还面临重大挑战，包括膳食和多组学数据存在偏倚且标准化不一致；非欧洲人群代表性不足，这限制了模型可泛化性并引发对健康公平性的担忧；模型可解释性有限；隐私和数据治理挑战；作为医疗器械的自适应软件监管路径尚不成熟；以及与","Frontiers in Nutrition","2026-09-28T00:00:00Z",{"impact":15,"substance":15,"depth":15,"authority":15,"freshness":15,"relevant":15,"comment":152},"该文为医学营养领域的AI综述，与三农、农业信息化、智慧农业无直接关联，不建议进入每日精选。",[154],{"name":149,"url":146},[21,156,22,157],"多组学","精准营养",[],{"status":26,"locked":14,"version":27,"method":28,"evidence":160},{"candidates":161,"agricultural_evidence":14,"reason":46},[162,168,174],{"code":32,"score":82,"supported":34,"title":163,"abstract":164,"tags":167},[110],[165,108,37,109,166,110,38],"深度学习","deep learning",[21],{"code":113,"score":169,"supported":34,"title":170,"abstract":171,"tags":173},3,[],[172],"数字孪生",[],{"code":175,"score":42,"supported":14,"title":176,"abstract":177,"tags":178},"F2",[],[],[156],[180,181],"AI 个性化营养","肠道微生物 营养干预","AI个性化营养-3858",4,"10.3389\u002Ffnut.2026.1974199",{"doi":184,"openalex_id":186,"authors":187,"venue":149,"cited_by_count":15,"oa_url":199,"card":200,"direction":205,"ingested_from":67},"W7214581081",[188,191,193,195,197],{"name":189,"orcid":190},"Ya Zhang","https:\u002F\u002Forcid.org\u002F0000-0003-0894-4922",{"name":192,"orcid":9},"Meijing Feng",{"name":194,"orcid":9},"Qiuyan Xue",{"name":196,"orcid":9},"Rui Ma",{"name":198,"orcid":9},"Fangming Liu","https:\u002F\u002Fwww.frontiersin.org\u002Fjournals\u002Fnutrition\u002Farticles\u002F10.3389\u002Ffnut.2026.1974199\u002Fpdf",{"tldr":201,"method":202,"finding":203,"direction":65,"opportunity":204},"综述AI在个性化营养干预中的应用，涵盖生物、技术、临床三层并评估证据强度。","综述多组学、机器学习、深度学习、LLM、知识图谱、数字孪生与CGM闭环系统。","T2DM\u002F糖尿病前期证据最成熟，AI个性化干预优于地中海饮食；肥胖、心血管、肿瘤等证据仍混合或初步。","可探索将AI个性化营养与农业食品生产、膳食推荐及可穿戴设备数据闭环结合，并关注数据偏差与公平性。","数字乡村与农业信息化","2026-09-30T23:30:41.633764Z",{"id":208,"title":209,"url":210,"summary":211,"summary_zh":212,"content":9,"source_name":213,"source_url":210,"published_at":214,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":215,"score_detail":216,"sources":219,"tags":221,"primary_topic":224,"secondary_topics":230,"classification":232,"search_phrases":248,"slug":251,"view_count":15,"doi":252,"paper":253,"created_at":267},5592,"Evaluation of Concept-Rank Collapse Audits for Computer Vision Agriculture Disease Classifiers Under Low-Label Regimes: An Interpretability Capacity Bound for Smallholder Fungicide Selectivity","https:\u002F\u002Fdoi.org\u002F10.5281\u002Fzenodo.23545193","Concept-based interpretability methods for plant-disease classifiers are typically evaluated by whether their concept activations correlate with human-annotated labels, but this criterion is silent on a prior question: whether the classifier's internal representation retains enough concept-rank structure for any audit to be informative. the present analysis formalises that prior question as an interpretability capacity bound. A classifier is modelled as a concept-rank operator whose singular values govern how much concept-level signal survives the projection from pixel space to class logits, and a collapse audit is defined as the ratio of retained concept-rank energy to total concept energy. Under a low-label regime, where concept annotations are scarce and the audit must be estimated from few labelled examples, the bound predicts that audit reliability degrades as a function of the concept-rank spectrum and the label budget. The analysis is a theoretical and methodological contribution: it derives the bound, states the conditions under which a collapse audit is identifiable, and characterises the operating characteristics of the audit as a decision procedure under a declared variance structure (between-cluster SD = 6.0, residual SD = 4.5, intraclass correlation = 0.64). Power was estimated across 2000 design cells for cluster counts of 8, 16, 24 and 32 and effect sizes of 3, 5 and 8 percentage points; 52 clusters were required for 5 percentage points at 80% power. The bound reframes concept audits as capacity tests rather than fidelity tests, with consequences for how smallholder fungicide-selectivity decisions should be supported by interpretable disease classifiers.","面向植物病害分类器的基于概念的可解释性方法，通常以其概念激活是否与人工标注标签相关作为评价标准，但该标准对一个更先前的问题保持沉默：分类器的内部表征是否保留了足够的概念秩结构，从而使任何审计具有信息量。本分析将该先前问题形式化为一个可解释性容量上界。分类器被建模为概念秩算子，其奇异值支配着从像素空间到类别logits的投影中概念级信号得以保留的程度，并将坍缩审计定义为保留的概念秩能量与总概念能量之比。在低标签情形下，概念标注稀缺且审计必须由少量标注样本估计，该上界预测审计可靠性会随概念秩谱和标签预算的函数而退化。本分析是一项理论和方法论贡献：推导出该上界，陈述坍缩审计可识别的条件，并在声明的方差结构（簇间SD = 6.0，残差SD = 4.5，组内相关系数 = 0.64）下刻画审计作为决策程序的操作特征。在2000个设计单元中，针对簇数为8、16、24和32以及效应量为3、5和8个百分点的情形估计了功效；在80%功效下达到5个百分点需要52个簇。该上界将概念审计重新框定为容量检验而非保真度检验，并对可解释病害分类器应如何支持小农杀菌剂选择性决策具有启示意义。","Zenodo (CERN European Organization for Nuclear Research)","2026-10-21T00:00:00Z",69,{"impact":217,"substance":83,"depth":82,"authority":105,"freshness":13,"relevant":42,"comment":218},8,"理论方法论文，提出可解释性容量上界并给出审计功效分析，对智慧农业病害识别可信度有参考价值，但属小众方法学研究、缺乏田间实证，暂不宜进入每日精选。",[220],{"name":213,"url":210},[91,21,222,93,223],"小农户","可解释性",{"code":225,"slug":226,"name":227,"path":228,"parent_code":229},"G2","farm-services","社会化服务与规模经营","\u002Ftopic\u002Frural-services\u002Ffarm-services","G",[231],{"code":32,"slug":95,"name":96,"path":97,"parent_code":98},{"status":101,"locked":14,"version":27,"method":28,"evidence":233},{"candidates":234,"agricultural_evidence":34},[235,240,244],{"code":225,"score":105,"supported":34,"title":236,"abstract":238,"tags":239},[237],"smallholder",[237],[222],{"code":32,"score":13,"supported":34,"title":241,"abstract":242,"tags":243},[38],[],[21,93],{"code":113,"score":42,"supported":14,"title":245,"abstract":246,"tags":247},[],[],[91],[249,250],"农业人工智能 植物病害识别 可解释性 智慧农业","农业人工智能 植物病害识别","农业人工智能植物病害识别可解释性智慧农业-5592","10.5281\u002Fzenodo.23545193",{"doi":252,"openalex_id":254,"authors":255,"venue":213,"cited_by_count":15,"oa_url":210,"card":262,"direction":65,"ingested_from":67},"W7223137323",[256,258,260],{"name":257,"orcid":9},"Akshay Ramgoolam",{"name":259,"orcid":9},"Priya Bundhoo Appadoo",{"name":261,"orcid":9},"Jean-Noël LEspérance",{"tldr":263,"method":264,"finding":265,"direction":65,"opportunity":266},"提出可解释性容量上界，评估低标签下概念秩崩溃审计的可靠性。","将分类器建模为概念秩算子，推导容量上界并做2000格功效模拟。","审计可靠性随概念秩谱与标签预算下降；5个百分点效应需52个聚类达80%功效。","可将容量上界用于小农户杀菌剂选择性决策的可解释模型审计与标签预算设计。","2026-10-22T11:21:10.425290Z",{"id":269,"title":270,"url":271,"summary":272,"summary_zh":273,"content":9,"source_name":213,"source_url":271,"published_at":214,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":215,"score_detail":274,"sources":277,"tags":279,"primary_topic":283,"secondary_topics":289,"classification":291,"search_phrases":316,"slug":319,"view_count":15,"doi":320,"paper":321,"created_at":333},5591,"Hierarchical Conformal Risk Control for Artificial Intelligence Agriculture Irrigation Scheduling Under Unmodeled Infiltration: A Computer Safety-Guarantee Evaluation","https:\u002F\u002Fdoi.org\u002F10.5281\u002Fzenodo.23543665","Irrigation scheduling algorithms that rely on soil-moisture forecasts are vulnerable to unmodeled infiltration, the process by which water bypasses the root zone through preferential pathways that resistance-based soil models cannot represent. The framework is evaluated through a replication-based design study that characterises the operating characteristics of the before-after-control-impact estimator used to audit scheduler performance, under a declared variance structure with between-cluster standard deviation 6.0, residual standard deviation 4.5 and intraclass correlation 0.64. Power was estimated across 2000 design cells spanning cluster counts of 8, 16, 24 and 32 and effect sizes of 3, 5 and 8 percentage points. The analysis indicates that 54 clusters are required to detect a 5-percentage-point effect at 80% power, and that the minimum detectable effect declines steeply as cluster count rises from 8 to 32. The contribution is a formal safety guarantee for irrigation decision support that remains valid when the infiltration model is wrong, together with an audit design that makes the guarantee empirically checkable.","依赖土壤湿度预报的灌溉调度算法容易受到未建模入渗的影响，即水分通过优先路径绕过根区的过程，而基于阻力的土壤模型无法表征这一过程。该框架通过一项基于重复的设计研究进行评估，该研究刻画了用于审计调度器性能的前后对照影响估计量的运行特性，其所依据的声明方差结构为：组间标准差6.0、残差标准差4.5、组内相关系数0.64。功效估计覆盖2000个设计单元，涵盖聚类数8、16、24和32以及效应量3、5和8个百分点。分析表明，在80%功效下检测5个百分点效应需要54个聚类，且最小可检测效应随聚类数从8增至32而急剧下降。本文的贡献在于为灌溉决策支持提供了一种正式的安全保证，该保证在入渗模型错误时仍然有效，同时提出了一种使该保证可经实证检验的审计设计。",{"impact":217,"substance":275,"depth":82,"authority":105,"freshness":86,"relevant":42,"comment":276},22,"该研究提出在入渗模型失配条件下仍有效的灌溉决策安全保证框架，并给出可审计的试验设计，方法新颖、结论可靠，但属细分方法学研究，产业影响有限，适合作为专业精选而非头条。",[278],{"name":213,"url":271},[91,21,280,281,282],"灌溉决策","安全保证","试验设计",{"code":284,"slug":285,"name":286,"path":287,"parent_code":288},"H3","irrigation","水资源与灌溉","\u002Ftopic\u002Fenvironment\u002Firrigation","H",[290],{"code":32,"slug":95,"name":96,"path":97,"parent_code":98},{"status":101,"locked":14,"version":27,"method":28,"evidence":292},{"candidates":293,"agricultural_evidence":34},[294,300,305,312],{"code":284,"score":295,"supported":34,"title":296,"abstract":297,"tags":299},15,[285],[298,285],"灌溉",[280],{"code":32,"score":105,"supported":34,"title":301,"abstract":302,"tags":304},[110],[303],"decision support",[21],{"code":306,"score":70,"supported":34,"title":307,"abstract":308,"tags":311},"H2",[],[309,310],"土壤","soil",[],{"code":113,"score":42,"supported":14,"title":313,"abstract":314,"tags":315},[],[],[91],[317,318],"灌溉调度 安全保证","农业人工智能 安全保证 智慧农业 灌溉决策","灌溉调度安全保证-5591","10.5281\u002Fzenodo.23543665",{"doi":320,"openalex_id":322,"authors":323,"venue":213,"cited_by_count":15,"oa_url":271,"card":328,"direction":65,"ingested_from":67},"W7223131205",[324,326],{"name":325,"orcid":9},"Nassima Benkhelifa",{"name":327,"orcid":9},"Yacine Aït Ahmed",{"tldr":329,"method":330,"finding":331,"direction":65,"opportunity":332},"提出分层保形风险控制，为未建模入渗下的灌溉调度提供安全保证。","分层保形风险控制与基于重复设计的BACI审计，2000个设计单元功效分析。","检测5个百分点效应需54个集群，最小可检测效应随集群数增加陡降。","可将保形风险控制扩展到其他未建模扰动下的农业决策安全保证与审计设计。","2026-10-22T11:21:10.283561Z",{"id":335,"title":336,"url":337,"summary":338,"summary_zh":339,"content":9,"source_name":213,"source_url":337,"published_at":214,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":340,"score_detail":341,"sources":343,"tags":345,"primary_topic":348,"secondary_topics":349,"classification":351,"search_phrases":366,"slug":369,"view_count":15,"doi":370,"paper":371,"created_at":385},5590,"Evaluation of Concept-Probe Temporal-Stability Audits for Computer Vision Agriculture Disease Classifiers Under Seasonal Phenology Drift: An Interpretability Decay Bound for Smallholder Fungicide Timing","https:\u002F\u002Fdoi.org\u002F10.5281\u002Fzenodo.23545234","Concept-probe audits are increasingly used to certify that a crop-disease classifier's internal representations track the pathological concepts a smallholder would act on, but a probe certified at one point in the season may silently decouple from those concepts as phenology shifts the visual statistics of the field. the present analysis treats that decoupling as a formal object. It defines a temporal-stability audit as a sequence of concept-probe accuracies measured across defined phenological windows, and it derives an interpretability decay bound: the rate at which probe accuracy can fall before the fungicide-timing decision it is meant to support becomes unreliable. The bound is expressed as a function of the probe's initial margin, the drift in concept prevalence between windows, and the classifier's sensitivity to the drifted features, and it yields a certified audit interval rather than a single-point certificate. The paper then characterises the operating characteristics of the audit design under a declared variance structure, using a replication-based evaluation of the before-after-control-impact estimator with between-cluster SD = 6.0, residual SD = 4.5 and intraclass correlation = 0.64. Power was estimated across 2000 design cells over cluster counts of 8, 16, 24 and 32 and effect sizes of 3, 5 and 8 percentage points; 54 clusters were required for 5 percentage points at 80% power. The contribution is a decision-theoretic account of when a concept-probe certificate expires, and a design rule for scheduling re-audits across the season.","概念探针审计（concept-probe audit）日益被用于验证作物病害分类器的内部表征是否追踪小农户会据以采取行动的病理学概念，但在生长季某一时点获得认证的探针，可能随着物候变化改变田间的视觉统计特征而悄然与这些概念脱钩。本分析将这种脱钩视为一个形式化对象。它将时间稳定性审计（temporal-stability audit）定义为在既定物候窗口内测量的一系列概念探针准确率，并推导出一个可解释性衰减界限（interpretability decay bound）：即探针准确率在其所支持的杀菌剂施用时机决策变得不可靠之前所能下降的速率。该界限表示为探针初始裕度、窗口间概念流行率漂移以及分类器对漂移特征敏感度的函数，并给出一个经认证的审计区间而非单点认证。随后，本文在声明的方差结构下刻画该审计设计的运行特征，采用基于重复的前后对照影响估计量评估，其中组间标准差 = 6.0、残差标准差 = 4.5、组内相关系数 = 0.64。功效在 2000 个设计单元上估计，聚类数取 8、16、24 和 32，效应量取 3、5 和 8 个百分点；在 80% 功效下达到 5 个百分点需要 54 个聚类。本文的贡献在于从决策论角度阐明概念探针认证何时失效，并给出在整个生长季内安排重新审计的设计规则。",70,{"impact":217,"substance":275,"depth":82,"authority":105,"freshness":13,"relevant":42,"comment":342},"该论文提出概念探针审计的时序稳定性框架与可解释性衰减界，方法新颖、统计设计扎实，但属小众方法论研究，产业影响有限，适合作为专业精选而非头条。",[344],{"name":213,"url":337},[91,21,222,346,347],"作物病害识别","模型审计",{"code":225,"slug":226,"name":227,"path":228,"parent_code":229},[350],{"code":32,"slug":95,"name":96,"path":97,"parent_code":98},{"status":101,"locked":14,"version":27,"method":28,"evidence":352},{"candidates":353,"agricultural_evidence":34},[354,358,362],{"code":225,"score":295,"supported":34,"title":355,"abstract":356,"tags":357},[237],[222,237],[222],{"code":32,"score":13,"supported":34,"title":359,"abstract":360,"tags":361},[38],[],[346,21],{"code":113,"score":42,"supported":14,"title":363,"abstract":364,"tags":365},[],[],[91],[367,368],"概念探针 时序稳定性 作物病害分类器","杀菌剂 施药时机 可解释性衰减","概念探针时序稳定性作物病害分类器-5590","10.5281\u002Fzenodo.23545234",{"doi":370,"openalex_id":372,"authors":373,"venue":213,"cited_by_count":15,"oa_url":337,"card":380,"direction":65,"ingested_from":67},"W7223088030",[374,376,378],{"name":375,"orcid":9},"Abdi-Basid Idriss Farah",{"name":377,"orcid":9},"Amina Abdillahi Guedi",{"name":379,"orcid":9},"Moussa Omar Youssouf",{"tldr":381,"method":382,"finding":383,"direction":65,"opportunity":384},"提出概念探针时间稳定性审计与可解释性衰减界，判断何时需重新审计。","定义跨物候窗口探针精度序列，推导衰减界，用BACI复制评估统计功效。","5个百分点效应需54个聚类达80%功效，审计证书会随物候漂移过期。","可将衰减界与再审计调度规则嵌入小农户杀菌剂时机决策支持系统，并验证跨作物泛化。","2026-10-22T11:21:08.978348Z",{"id":387,"title":388,"url":389,"summary":390,"summary_zh":391,"content":9,"source_name":213,"source_url":389,"published_at":214,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":392,"score_detail":393,"sources":395,"tags":397,"primary_topic":400,"secondary_topics":401,"classification":402,"search_phrases":418,"slug":421,"view_count":15,"doi":422,"paper":423,"created_at":433},5589,"Conformal-Risk Evaluation for Artificial Intelligence Agriculture Spray-Nozzle Fault Detection Under Label-Delay Feedback: A Computer Coverage Audit","https:\u002F\u002Fdoi.org\u002F10.5281\u002Fzenodo.23544107","Spray-nozzle fault detection in precision agriculture is a classification problem in which the label that confirms a fault arrives after the spray pass, so the detector must act on unlabelled evidence and its error rate must be certified before the label exists. Conformal prediction offers distribution-free coverage guarantees, but the guarantee is stated for exchangeable calibration and test data, an assumption that label-delayed feedback violates because the calibration set is always older than the deployment set. the present analysis formalises label-delay feedback as a distribution shift in the conformity score and derives a coverage bound for split-conformal fault detectors under that shift. The bound separates the nominal miscoverage rate into an exchangeable component and a delay-induced component that scales with the total variation distance between the calibration and deployment score distributions. A replication-based evaluation of the ['Intercept', 'exposure', 'group', 'severity'] estimator was conducted: 600 data sets of N = 500 were drawn from a declared data-generating process (intercept 12.0, exposure 0.42, group -3.1, severity 1.35, error SD 4.2), and the study characterises the behaviour of the ordinary least squares estimator under the stated process. The estimator recovered the declared coefficients with estimated bias 0.01122 for the intercept and -0.0002 for exposure, root mean squared error 0.9217 and 0.01614 respectively, and empirical coverage of the nominal 95% intervals of 1.0 for both parameters across 600 replications. The coverage audit shows that the delay-induced term, not the exchangeable term, dominates the risk of a conformal spray-nozzle detector, and that the audit procedure can certify a detector before field labels return.","在精准农业中，喷雾喷嘴故障检测是一个分类问题，其中确认故障的标签在喷雾作业结束后才到达，因此检测器必须基于未标记的证据采取行动，并且其错误率必须在标签存在之前得到认证。保形预测（conformal prediction）提供了无分布覆盖保证，但该保证是针对可交换的校准数据和测试数据而言的，而标签延迟反馈违反了这一假设，因为校准集总是比部署集更旧。本分析将标签延迟反馈形式化为一致性分数中的分布偏移，并推导了在该偏移下分裂保形故障检测器的覆盖界。该界将名义错误覆盖 rate 分解为可交换分量和延迟诱导分量，后者随校准与部署分数分布之间的全变差距离而缩放。对 ['Intercept', 'exposure', 'group', 'severity'] 估计器进行了基于重复的评估：从声明的数据生成过程（截距 12.0，暴露 0.42，分组 -3.1，严重程度 1.35，误差标准差 4.2）中抽取了 600 个 N = 500 的数据集，研究刻画了普通最小二乘估计器在所述过程下的行为。该估计器恢复了声明的系数，截距的估计偏差为 0.01122，暴露为 -0.0002，均方根误差分别为 0.9217 和 0.01614，两个参数在 600 次重复中名义 95% 区间的经验覆盖率为 1.0。覆盖审计表明，延迟诱导项而非可交换项主导了保形喷雾喷嘴检测器的风险，并且该审计程序可以在田间标签返回之前对检测器进行认证。",62,{"impact":217,"substance":82,"depth":84,"authority":86,"freshness":13,"relevant":42,"comment":394},"论文提出标签延迟下共形预测覆盖界，方法有新意但属预印本、应用面窄，可作为专业补充而非每日精选头条。",[396],{"name":213,"url":389},[91,21,398,399],"精准施药","病虫害监测",{"code":32,"slug":95,"name":96,"path":97,"parent_code":98},[],{"status":101,"locked":14,"version":27,"method":28,"evidence":403},{"candidates":404,"agricultural_evidence":34},[405,409,413],{"code":32,"score":86,"supported":34,"title":406,"abstract":407,"tags":408},[110],[],[21],{"code":113,"score":33,"supported":34,"title":410,"abstract":411,"tags":412},[],[116,117],[91],{"code":414,"score":42,"supported":14,"title":415,"abstract":416,"tags":417},"F4",[],[],[398],[419,420],"农业人工智能 病虫害监测 智慧农业 精准施药","农业人工智能 病虫害监测","农业人工智能病虫害监测智慧农业精准施药-5589","10.5281\u002Fzenodo.23544107",{"doi":422,"openalex_id":424,"authors":425,"venue":213,"cited_by_count":15,"oa_url":389,"card":428,"direction":65,"ingested_from":67},"W7223116087",[426],{"name":427,"orcid":9},"Achieng' Atim Okello",{"tldr":429,"method":430,"finding":431,"direction":65,"opportunity":432},"将标签延迟反馈建模为共形分数分布偏移，推导覆盖率界并审计喷雾喷嘴故障检测器。","分裂共形预测、总变差距离、600次重复模拟（N=500）评估OLS估计。","延迟诱导项而非可交换项主导检测器风险，审计可在田间标签返回前认证检测器。","可将该覆盖率审计框架扩展到其他标签延迟的农业传感器故障检测，并研究延迟分布估计误差对认证的影响。","2026-10-22T11:21:02.845718Z"]