[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"item-2281":3,"related-2281":98},{"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":23,"tags":25,"primary_topic":31,"secondary_topics":37,"classification":38,"search_phrases":63,"slug":66,"view_count":53,"doi":67,"paper":68,"created_at":97},2281,"Artificial intelligence-enabled aquacultural engineering systems: A design-oriented review of sensing, modelling, and deployment architectures","https:\u002F\u002Fdoi.org\u002F10.1016\u002Fj.compag.2026.112353","Artificial intelligence-enabled aquacultural engineering systems: A design-oriented review of sensing, modelling, and deployment architectures。Computers and Electronics in Agriculture","人工智能赋能的水产养殖工程系统：面向设计的传感、建模与部署架构综述。农业中的计算机与电子学",null,"Computers and Electronics in Agriculture","2026-09-11T00:00:00Z","论文",10,false,78,{"impact":17,"substance":18,"depth":17,"authority":19,"freshness":20,"relevant":21,"comment":22},18,20,14,8,1,"核心期刊发表的水产养殖AI工程系统设计综述，涵盖感知、建模与部署架构，对智慧渔业技术路线有参考价值，但属综述类论文，产业落地影响有限。",[24],{"name":10,"url":6},[26,27,28,29,30],"智慧农业","农业人工智能","水产养殖","智能传感","数字渔业",{"code":32,"slug":33,"name":34,"path":35,"parent_code":36},"I2","agricultural-ai","农业AI与大模型","\u002Ftopic\u002Fdigital-agriculture\u002Fagricultural-ai","I",[],{"status":39,"locked":14,"version":40,"method":41,"evidence":42},"classified","2026-10-22.2","rules",{"candidates":43,"agricultural_evidence":46},[44,51,58],{"code":32,"score":45,"supported":46,"title":47,"abstract":49,"tags":50},12,true,[48],"artificial intelligence",[48],[27],{"code":52,"score":53,"supported":46,"title":54,"abstract":55,"tags":57},"D",4,[],[56],"水产",[28],{"code":59,"score":21,"supported":14,"title":60,"abstract":61,"tags":62},"I1",[],[],[26],[64,65],"农业人工智能 数字渔业 智慧农业 智能传感","农业人工智能 数字渔业","农业人工智能数字渔业智慧农业智能传感-2281","10.1016\u002Fj.compag.2026.112353",{"doi":67,"openalex_id":69,"authors":70,"venue":10,"cited_by_count":89,"oa_url":6,"card":90,"direction":94,"ingested_from":96},"W7212232512",[71,74,77,80,83,86],{"name":72,"orcid":73},"V. Ravi Sankar","https:\u002F\u002Forcid.org\u002F0000-0002-8580-2920",{"name":75,"orcid":76},"Alzayat Saleh","https:\u002F\u002Forcid.org\u002F0000-0001-6973-019X",{"name":78,"orcid":79},"Armin Ehrampoosh","https:\u002F\u002Forcid.org\u002F0000-0002-5482-3454",{"name":81,"orcid":82},"Phoebe Arbon","https:\u002F\u002Forcid.org\u002F0000-0003-0171-6302",{"name":84,"orcid":85},"Dean R. Jerry","https:\u002F\u002Forcid.org\u002F0000-0003-3735-1798",{"name":87,"orcid":88},"Mostafa Rahimi Azghadi","https:\u002F\u002Forcid.org\u002F0000-0001-7975-3985",0,{"tldr":91,"method":92,"finding":93,"direction":94,"opportunity":95},"综述AI赋能水产养殖工程系统，从感知、建模到部署架构的设计视角。","设计导向综述，梳理传感器、AI建模与部署架构。","提出面向水产养殖的AI工程系统设计框架与架构分类。","智慧农业 \u002F 农业物联网","可针对水产养殖场景的轻量化边缘AI与多模态传感融合部署开展实证研究。","openalex","2026-09-13T23:30:01.790828Z",{"total":99,"page":21,"page_size":99,"items":100},6,[101,169,237,299,363,415],{"id":102,"title":103,"url":104,"summary":105,"summary_zh":106,"content":9,"source_name":107,"source_url":104,"published_at":108,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":109,"score_detail":110,"sources":115,"tags":117,"primary_topic":120,"secondary_topics":121,"classification":122,"search_phrases":142,"slug":145,"view_count":146,"doi":147,"paper":148,"created_at":168},2913,"A quantum-inspired multi-objective learning framework for real-time sustainable aquaculture water quality prediction","https:\u002F\u002Fdoi.org\u002F10.3389\u002Ffsufs.2026.1827991","Management of sustainable aquaculture necessitates accurate prediction of water quality parameters, since environmental variability is an important factor in determining aquatic productivity and ecological stability. Traditional machine learning systems are usually limited by inefficient parameter optimization, feature overlap, and limited adaptability to non-linear and time-varying environmental trends. To address these issues, a new concept is proposed, which is called Quantum-Inspired Aquaculture Optimization Network (Q-AQUAOptNet). To improve the process of feature selection, hyperparameter optimization, and model stability before temporal prediction, the proposed architecture combines sustainability index modeling and quantum-inspired multi-objective evolutionary optimization. The optimization approach balances exploration and exploitation, enhancing the learning ability of the prediction model. Implementation of the framework was done through Python-based simulation tools in preprocessing, optimization, and performance evaluation. The created system had a Root Mean Square Error of 0.24, a Mean Absolute Error of 0.21, and an R 2 of 0.97, which means that it has effective predictive power and a great ability to explain the variance. Integrating evolutionary optimization and temporal learning improves generalization performance and reduces prediction uncertainty. The findings demonstrate that Q-AQUAOptNet provides effective predictive performance within a simulation-based framework and shows potential as a sustainability-oriented intelligent water-quality monitoring and decision support system for aquaculture applications.","可持续水产养殖管理需要准确预测水质参数，因为环境变异性是决定水生生产力和生态稳定性的重要因素。传统机器学习系统通常受限于参数优化效率低、特征重叠以及对非线性和时变环境趋势的适应能力有限。为解决这些问题，提出了一种新概念，称为量子启发水产养殖优化网络（Quantum-Inspired Aquaculture Optimization Network，Q-AQUAOptNet）。为在时间预测之前改进特征选择、超参数优化和模型稳定性，所提出的架构结合了可持续性指数建模和量子启发多目标进化优化。该优化方法平衡了探索与利用，增强了预测模型的学习能力。该框架的实现通过基于Python的仿真工具完成，涵盖预处理、优化和性能评估。所构建系统的均方根误差为0.24，平均绝对误差为0.21，R²为0.97，这意味着其具有有效的预测能力和较强的方差解释能力。将进化优化与时间学习相结合，提高了泛化性能并降低了预测不确定性。研究结果表明，Q-AQUAOptNet在基于仿真的框架内提供了有效的预测性能，并显示出作为面向可持续性的智能水质监测与决策支持系统应用于水产养殖的潜力。","Frontiers in Sustainable Food Systems","2026-09-18T00:00:00Z",71,{"impact":45,"substance":18,"depth":111,"authority":112,"freshness":113,"relevant":21,"comment":114},17,13,9,"提出量子启发多目标优化网络用于水产养殖水质预测，仿真指标较优，属智慧渔业细分领域的方法学进展，但尚处仿真阶段、缺乏真实场景验证。",[116],{"name":107,"url":104},[26,27,28,118,119],"多目标优化","水质预测",{"code":32,"slug":33,"name":34,"path":35,"parent_code":36},[],{"status":39,"locked":14,"version":40,"method":41,"evidence":123},{"candidates":124,"agricultural_evidence":46},[125,131,138],{"code":52,"score":126,"supported":46,"title":127,"abstract":129,"tags":130},15,[128],"aquaculture",[56,128],[28],{"code":32,"score":13,"supported":46,"title":132,"abstract":133,"tags":137},[],[134,135,136],"机器学习","machine learning","decision support",[27],{"code":59,"score":21,"supported":14,"title":139,"abstract":140,"tags":141},[],[],[26],[143,144],"Q-AQUAOptNet 水产养殖 水质预测","农业人工智能 多目标优化 智慧农业 水产养殖","Q-AQUAOptNet水产养殖水质预测-2913",3,"10.3389\u002Ffsufs.2026.1827991",{"doi":147,"openalex_id":149,"authors":150,"venue":107,"cited_by_count":89,"oa_url":104,"card":163,"direction":94,"ingested_from":96},"W7213532503",[151,154,157,159,161],{"name":152,"orcid":153},"Abdel‐Haleem Abdel‐Aty","https:\u002F\u002Forcid.org\u002F0000-0002-6763-2569",{"name":155,"orcid":156},"Ali Jaber Almalki","https:\u002F\u002Forcid.org\u002F0009-0004-1359-2612",{"name":158,"orcid":9},"Sara A. Ghorashi",{"name":160,"orcid":9},"Betty Wan Niu Voon",{"name":162,"orcid":9},"Mohamed Hafez",{"tldr":164,"method":165,"finding":166,"direction":94,"opportunity":167},"提出量子启发多目标优化网络Q-AQUAOptNet，实现水产养殖水质实时预测。","量子启发多目标进化优化结合时间预测与可持续指数建模，Python仿真。","RMSE 0.24、MAE 0.21、R² 0.97，预测精度高且泛化好。","可探索真实养殖场部署与多源传感器融合，验证量子启发优化在边缘端的实时性。","2026-09-19T23:30:07.985707Z",{"id":170,"title":171,"url":172,"summary":173,"summary_zh":174,"content":9,"source_name":175,"source_url":172,"published_at":176,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":15,"score_detail":177,"sources":180,"tags":182,"primary_topic":185,"secondary_topics":190,"classification":192,"search_phrases":213,"slug":216,"view_count":53,"doi":217,"paper":218,"created_at":236},2020,"Decision support in recirculating aquaculture systems (RAS): A case study of a human–AI interface in prawn hatchery operation","https:\u002F\u002Fdoi.org\u002F10.1016\u002Fj.atech.2026.102554","The need for sustainable and responsible production in aquaculture calls for innovative implementation of recirculating aquaculture systems (RAS), which are extraordinarily complex, requiring the integration of various fields of science and technology to reach the desired productivity. In this case study, we report a four-month observation using a large language model (LLM) that – collaboratively with human expertise – analyzed and resolved complex challenges in a Macrobrachium rosenbergii RAS hatchery. Unacceptable larval and post-larval mortality prompted the integration into hatchery management of an AI decision-support system as a strategic management partner, enabling exploration across chemical, biological, physical, engineering, and behavioral domains. Key interventions suggested by the AI agent included mineral balance recalibration, microbial load diagnostics, behavioral pattern decoding, and lighting and flow engineering. Outcomes were evaluated in terms of a reduction in larval mortality and improved rates of larval metamorphosis to post larvae. Central to the process was the presence of a guiding human entity, steering AI's analytical power through deliberate questioning and contextual framing. This case study suggests that AI has the potential to improve intensive aquaculture systems. However, the tendency of AI agents to oversimplify complex systems requires the direction and guidance of a human expert to lead AI-human conversations. The adoption of LLMs in RAS-based aquaculture, bridging the gaps between raw data and actionable insights, has the potential to drive both the efficiency and the long-term sustainability of the aquaculture industry.","水产养殖可持续和负责任生产的需要，要求创新性地实施循环水养殖系统（RAS），该系统极为复杂，需要整合各种科学和技术领域以实现理想的生产力。在本案例研究中，我们报告了一项为期四个月的观察，使用大型语言模型（LLM）与人类专业知识协作，分析和解决了罗氏沼虾（Macrobrachium rosenbergii）RAS孵化场中的复杂挑战。不可接受的幼体和后期幼体死亡率促使将AI决策支持系统作为战略管理伙伴纳入孵化场管理，从而能够在化学、生物、物理、工程和行为领域进行探索。AI代理建议的关键干预措施包括矿物质平衡重新校准、微生物负荷诊断、行为模式解码以及光照和水流工程。结果通过幼体死亡率的降低和幼体变态为后期幼体的比率提高来评估。该过程的核心是有一个指导性的人类实体，通过有意的提问和情境构建来引导AI的分析能力。本案例研究表明，AI有潜力改善集约化水产养殖系统。然而，AI代理倾向于过度简化复杂系统，需要人类专家的指导和引导来主导AI与人类的对话。在基于RAS的水产养殖中采用LLM，弥合原始数据与可操作见解之间的差距，有潜力推动水产养殖行业的效率和长期可持续性。","Smart Agricultural Technology","2026-09-07T00:00:00Z",{"impact":17,"substance":178,"depth":17,"authority":112,"freshness":20,"relevant":21,"comment":179},21,"核心期刊发表的LLM辅助RAS对虾育苗决策案例，方法新颖、结论有实证支撑，对智慧水产养殖具参考价值，值得进入每日精选。",[181],{"name":175,"url":172},[26,27,183,28,184],"大语言模型","循环水养殖",{"code":186,"slug":187,"name":188,"path":189,"parent_code":52},"D1","pond-factory-aquaculture","池塘与工厂化养殖","\u002Ftopic\u002Faquaculture\u002Fpond-factory-aquaculture",[191],{"code":32,"slug":33,"name":34,"path":35,"parent_code":36},{"status":39,"locked":14,"version":40,"method":41,"evidence":193},{"candidates":194,"agricultural_evidence":46},[195,199,204,209],{"code":52,"score":126,"supported":46,"title":196,"abstract":197,"tags":198},[128],[56,128],[28],{"code":186,"score":19,"supported":46,"title":200,"abstract":202,"tags":203},[201],"recirculating aquaculture",[184,201],[],{"code":32,"score":112,"supported":46,"title":205,"abstract":206,"tags":208},[136],[207],"large language model",[27,183],{"code":59,"score":21,"supported":14,"title":210,"abstract":211,"tags":212},[],[],[26],[214,215],"农业人工智能 大语言模型 循环水养殖 智慧农业","农业人工智能 大语言模型","农业人工智能大语言模型循环水养殖智慧农业-2020","10.1016\u002Fj.atech.2026.102554",{"doi":217,"openalex_id":219,"authors":220,"venue":175,"cited_by_count":89,"oa_url":229,"card":230,"direction":94,"ingested_from":96},"W7167492172",[221,224,226],{"name":222,"orcid":223},"Shai Avraham Shaked","https:\u002F\u002Forcid.org\u002F0000-0003-1995-6419",{"name":225,"orcid":9},"Assaf Shechter",{"name":227,"orcid":228},"Amir Sagi","https:\u002F\u002Forcid.org\u002F0000-0002-4229-1059","https:\u002F\u002Fwww.sciencedirect.com\u002Fscience\u002Farticle\u002Fpii\u002FS2772375526007793\u002Fpdf",{"tldr":231,"method":232,"finding":233,"direction":234,"opportunity":235},"用大语言模型辅助人类专家，解决罗氏沼虾RAS育苗中幼体高死亡率问题。","四个月案例观察，LLM与人类专家协作分析化学、生物、工程等多域数据。","AI建议的矿物质平衡、微生物诊断等干预降低了幼体死亡率并提高变态率。","农业人工智能与决策模型","可研究LLM在RAS多参数耦合决策中的可解释性与人机协同机制，避免过度简化。","2026-09-10T23:30:03.364805Z",{"id":238,"title":239,"url":240,"summary":241,"summary_zh":242,"content":9,"source_name":243,"source_url":240,"published_at":244,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":245,"score_detail":246,"sources":248,"tags":250,"primary_topic":254,"secondary_topics":260,"classification":262,"search_phrases":279,"slug":282,"view_count":89,"doi":283,"paper":284,"created_at":298},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":20,"substance":178,"depth":17,"authority":45,"freshness":13,"relevant":21,"comment":247},"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。",[249],{"name":243,"url":240},[26,27,251,252,253],"小农户","植物病害识别","可解释性",{"code":255,"slug":256,"name":257,"path":258,"parent_code":259},"G2","farm-services","社会化服务与规模经营","\u002Ftopic\u002Frural-services\u002Ffarm-services","G",[261],{"code":32,"slug":33,"name":34,"path":35,"parent_code":36},{"status":39,"locked":14,"version":40,"method":41,"evidence":263},{"candidates":264,"agricultural_evidence":46},[265,270,275],{"code":255,"score":45,"supported":46,"title":266,"abstract":268,"tags":269},[267],"smallholder",[267],[251],{"code":32,"score":13,"supported":46,"title":271,"abstract":273,"tags":274},[272],"computer vision",[],[27,252],{"code":59,"score":21,"supported":14,"title":276,"abstract":277,"tags":278},[],[],[26],[280,281],"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":283,"openalex_id":285,"authors":286,"venue":243,"cited_by_count":89,"oa_url":240,"card":293,"direction":234,"ingested_from":96},"W7223137323",[287,289,291],{"name":288,"orcid":9},"Akshay Ramgoolam",{"name":290,"orcid":9},"Priya Bundhoo Appadoo",{"name":292,"orcid":9},"Jean-Noël LEspérance",{"tldr":294,"method":295,"finding":296,"direction":234,"opportunity":297},"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":300,"title":301,"url":302,"summary":303,"summary_zh":304,"content":9,"source_name":243,"source_url":302,"published_at":244,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":245,"score_detail":305,"sources":308,"tags":310,"primary_topic":314,"secondary_topics":320,"classification":322,"search_phrases":345,"slug":348,"view_count":89,"doi":349,"paper":350,"created_at":362},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":20,"substance":306,"depth":17,"authority":45,"freshness":113,"relevant":21,"comment":307},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。",[309],{"name":243,"url":302},[26,27,311,312,313],"灌溉决策","安全保证","试验设计",{"code":315,"slug":316,"name":317,"path":318,"parent_code":319},"H3","irrigation","水资源与灌溉","\u002Ftopic\u002Fenvironment\u002Firrigation","H",[321],{"code":32,"slug":33,"name":34,"path":35,"parent_code":36},{"status":39,"locked":14,"version":40,"method":41,"evidence":323},{"candidates":324,"agricultural_evidence":46},[325,330,334,341],{"code":315,"score":126,"supported":46,"title":326,"abstract":327,"tags":329},[316],[328,316],"灌溉",[311],{"code":32,"score":45,"supported":46,"title":331,"abstract":332,"tags":333},[48],[136],[27],{"code":335,"score":99,"supported":46,"title":336,"abstract":337,"tags":340},"H2",[],[338,339],"土壤","soil",[],{"code":59,"score":21,"supported":14,"title":342,"abstract":343,"tags":344},[],[],[26],[346,347],"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":349,"openalex_id":351,"authors":352,"venue":243,"cited_by_count":89,"oa_url":302,"card":357,"direction":234,"ingested_from":96},"W7223131205",[353,355],{"name":354,"orcid":9},"Nassima Benkhelifa",{"name":356,"orcid":9},"Yacine Aït Ahmed",{"tldr":358,"method":359,"finding":360,"direction":234,"opportunity":361},"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":364,"title":365,"url":366,"summary":367,"summary_zh":368,"content":9,"source_name":243,"source_url":366,"published_at":244,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":369,"score_detail":370,"sources":372,"tags":374,"primary_topic":377,"secondary_topics":378,"classification":380,"search_phrases":395,"slug":398,"view_count":89,"doi":399,"paper":400,"created_at":414},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":20,"substance":306,"depth":17,"authority":45,"freshness":13,"relevant":21,"comment":371},"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。",[373],{"name":243,"url":366},[26,27,251,375,376],"作物病害识别","模型审计",{"code":255,"slug":256,"name":257,"path":258,"parent_code":259},[379],{"code":32,"slug":33,"name":34,"path":35,"parent_code":36},{"status":39,"locked":14,"version":40,"method":41,"evidence":381},{"candidates":382,"agricultural_evidence":46},[383,387,391],{"code":255,"score":126,"supported":46,"title":384,"abstract":385,"tags":386},[267],[251,267],[251],{"code":32,"score":13,"supported":46,"title":388,"abstract":389,"tags":390},[272],[],[375,27],{"code":59,"score":21,"supported":14,"title":392,"abstract":393,"tags":394},[],[],[26],[396,397],"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":399,"openalex_id":401,"authors":402,"venue":243,"cited_by_count":89,"oa_url":366,"card":409,"direction":234,"ingested_from":96},"W7223088030",[403,405,407],{"name":404,"orcid":9},"Abdi-Basid Idriss Farah",{"name":406,"orcid":9},"Amina Abdillahi Guedi",{"name":408,"orcid":9},"Moussa Omar Youssouf",{"tldr":410,"method":411,"finding":412,"direction":234,"opportunity":413},"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":416,"title":417,"url":418,"summary":419,"summary_zh":420,"content":9,"source_name":243,"source_url":418,"published_at":244,"category":12,"cover_url":9,"hotness":13,"is_selected":14,"score":421,"score_detail":422,"sources":424,"tags":426,"primary_topic":429,"secondary_topics":430,"classification":431,"search_phrases":450,"slug":453,"view_count":89,"doi":454,"paper":455,"created_at":465},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":20,"substance":17,"depth":111,"authority":113,"freshness":13,"relevant":21,"comment":423},"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。",[425],{"name":243,"url":418},[26,27,427,428],"精准施药","病虫害监测",{"code":32,"slug":33,"name":34,"path":35,"parent_code":36},[],{"status":39,"locked":14,"version":40,"method":41,"evidence":432},{"candidates":433,"agricultural_evidence":46},[434,438,445],{"code":32,"score":113,"supported":46,"title":435,"abstract":436,"tags":437},[48],[],[27],{"code":59,"score":439,"supported":46,"title":440,"abstract":441,"tags":444},7,[],[442,443],"精准农业","precision agriculture",[26],{"code":446,"score":21,"supported":14,"title":447,"abstract":448,"tags":449},"F4",[],[],[427],[451,452],"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":454,"openalex_id":456,"authors":457,"venue":243,"cited_by_count":89,"oa_url":418,"card":460,"direction":234,"ingested_from":96},"W7223116087",[458],{"name":459,"orcid":9},"Achieng' Atim Okello",{"tldr":461,"method":462,"finding":463,"direction":234,"opportunity":464},"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"]