[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"item-432":3,"related-432":67},{"id":4,"title":5,"url":6,"summary":7,"summary_zh":8,"content":9,"source_name":10,"source_url":8,"published_at":11,"category":12,"cover_url":8,"hotness":13,"is_selected":14,"score":15,"score_detail":16,"sources":24,"tags":26,"primary_topic":31,"secondary_topics":37,"classification":38,"search_phrases":62,"slug":65,"view_count":22,"doi":8,"paper":8,"created_at":66},432,"中国热带农业科学院发布'树薯智汇1.0'木薯产业垂类大模型——热带农业人工智能创新联盟成立","http:\u002F\u002Fwww.hq.xinhua.org\u002F20260808\u002F0a4e336974634a66939cc0d35a15ade8\u002Fc.html","8月6日,中国热带农业科学院科技信息研究所在海口发布'树薯智汇1.0'大模型,这是覆盖全产业链的木薯产业垂类大模型。系统整合品种资料、病虫草害图谱、产业数据、行业标准和专家经验,设置产业行情、品种推荐、病虫草害识别、品质智能分析、繁育栽培、木薯科普、木薯厨房、行业标准八大核心功能模块,通过微信小程序等便捷入口提供给农户、农技人员、收购商、加工企业及科研人员。中国热科院信息所、海南农科院、广东农科院、云南农科院经信所等9家单位联合成立'木薯产业人工智能创新联盟'。",null,"木薯作为我国热带、亚热带地区重要的粮食作物和工业原料，广泛应用于生产淀粉、饲料、食品、生物能源等产品。广西、海南、广东、云南等地拥有较好的木薯产业基础，但在实际生产经营中，品种和技术信息较为分散、市场行情获取渠道有限、上下游供需对接效率不高等问题，仍在一定程度上制约产业提质增效。\n\n中国热带农业科学院科技信息研究所所长叶剑秋介绍，“树薯智汇1.0”大模型正是从这些具体问题出发，将分散的专业知识、产业数据和专家经验进行系统整合，面向木薯种植、收购、加工、科普和标准服务等全产业链场景，通过微信小程序等便捷入口，为农户、农技人员、收购商、加工企业及科研人员提供智能服务，让科研成果从“资料库”走向生产一线，让专业服务从“人找信息”逐步转向“信息找人”。\n\n对于想了解木薯安全食用和加工方法的普通用户，平台设置了木薯科普、木薯厨房等模块；针对企业和从业人员，平台还可提供原料验收、质量检测、储运规范和行业标准等方面的查询服务，形成覆盖木薯产业关键环节的核心功能。","新华网海南 2026-08-08","2026-08-08T00:00:00Z","报道",10,false,72,{"impact":17,"substance":18,"depth":19,"authority":20,"freshness":21,"relevant":22,"comment":23},18,20,15,13,6,1,"国家级科研机构发布木薯产业垂类大模型，面向全产业链提供智能服务，具有产业级影响，信息增量明确，但报道深度一般，时效性稍差。",[25],{"name":10,"url":6},[27,28,29,30],"智慧农业","木薯","热带农业","垂类大模型",{"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},[44,52,57],{"code":32,"score":45,"supported":46,"title":47,"abstract":50,"tags":51},19,true,[48,49],"农业人工智能","大模型",[49],[],{"code":53,"score":22,"supported":14,"title":54,"abstract":55,"tags":56},"A5",[],[],[28],{"code":58,"score":22,"supported":14,"title":59,"abstract":60,"tags":61},"I1",[],[],[27],[63,64],"垂类大模型 智慧农业 热带农业 木薯","垂类大模型 智慧农业","垂类大模型智慧农业热带农业木薯-432","2026-08-12T23:57:20.540629Z",{"total":21,"page":22,"page_size":21,"items":68},[69,153,225,316,378,436],{"id":70,"title":71,"url":72,"summary":73,"summary_zh":74,"content":8,"source_name":75,"source_url":72,"published_at":76,"category":77,"cover_url":8,"hotness":13,"is_selected":14,"score":78,"score_detail":79,"sources":84,"tags":86,"primary_topic":90,"secondary_topics":91,"classification":92,"search_phrases":114,"slug":117,"view_count":118,"doi":119,"paper":120,"created_at":152},4423,"A stage-weighted environmental risk and visual severity integration model for climate-smart management of cassava anthracnose in humid tropical systems","https:\u002F\u002Fdoi.org\u002F10.3389\u002Ffphgy.2026.1898224","Colletotrichum anthracnose is a persistent constraint on cassava root yield and planting material quality in humid tropical systems, where calendar-based management often leads to poorly timed or excessive fungicide use. This study developed and pilot-tested the Cassava Anthracnose Risk Index (CARI), a stage-sensitive, persistence-adjusted early-warning framework integrating low-cost environmental sensing with structured visual severity scoring for climate-smart, input-reduced disease management. Microclimate sensors were deployed at CSIR-CRI Fumesua, Ghana, across a 13-month rainfed crop cycle (March 2025 – April 2026) using the susceptible landrace Debo . A stage-weighted Environmental Risk Subscore (ERS; 0–600 points, ERS%) was constructed from six environmental components, moderated by a persistence factor requiring sustained infection-favorable conditions, and stratified across three growth stages. A Visual Symptom Subscore (VSS; 0–40) was derived from standardised weekly severity assessments of 15 sentinel plants, integrated into a composite CARI (0–100) with stage-specific alert thresholds. Across 73 valid Stage 3 monitoring days, CARI classified all observations as Low Risk and issued no Amber or Red alerts, including on 37 of 73 days with sustained nocturnal high-humidity conditions that would have triggered a precautionary alert under a simple threshold approach. In the single confirmed disease episode observed, an approximately one-day interval separated the highest environmental risk signal from confirmed symptom onset; this is not presented as a validated lag value, a preceding 25-day sensor outage means antecedent conditions cannot be characterised. CARI did not issue an intervention recommendation during this episode (11–24 February 2026) because concurrent environmental risk remained below the Amber threshold. Published decision support systems for Colletotrichum anthracnose in other pathosystems report fungicide reductions of 30–70% relative to calendar-based approaches; these figures are cited only for benchmarking and were not measured here, as no comparative fungicide trial was conducted. These results provide proof of concept for the technical operability of CARI and identify the dataset and analyses required for formal predictive evaluation in a planned multi-season study; they should not be interpreted as evidence of CARI’s predictive accuracy or input-reduction performance.","炭疽病（Colletotrichum anthracnose）是湿润热带地区木薯块根产量和种植材料质量的持续制约因素，当地基于日历的防治方式常导致杀菌剂施用时机不当或过量。本研究开发并初步测试了木薯炭疽病风险指数（Cassava Anthracnose Risk Index, CARI），这是一个阶段敏感、持续性校正的预警框架，将低成本环境传感与结构化视觉严重度评分相结合，服务于气候智慧型、减投入的病害管理。微气候传感器部署于加纳CSIR-CRI Fumesua试验站，覆盖13个月雨养作物周期（2025年3月–2026年4月），供试品种为感病地方品种Debo。研究构建了阶段加权环境风险子评分（Environmental Risk Subscore, ERS；0–600分，ERS%），由六个环境组分构成，并受持续性因子调节——该因子要求感染有利条件持续存在——同时按三个生长阶段进行分层。视觉症状子评分（Visual Symptom Subscore, VSS；0–40）来源于对15株哨兵植株的标准化每周严重度评估，并与ERS整合为综合CARI（0–100），设定阶段特异性警报阈值。在73个有效的第3阶段监测日中，CARI将所有观测归类为低风险，未发出琥珀色或红色警报，其中包括73天中的37天存在持续性夜间高湿条件——若采用简单阈值方法，这些条件本会触发预防性警报。在观察到的唯一一次确诊病害事件中，最高环境风险信号与确诊症状出现之间相隔约一天；此值不作为经验证的滞后值呈现，因为此前25天的传感器中断意味着无法表征先行条件。CARI在该事件期间（2026年2月11–24日）未发出干预建议，因为同期环境风险仍低于琥珀色阈值。其他病理系统中针对炭疽病（Colletotrichum anthracnose）已发表的决策支持系统报告，相对于基于日历的方法，杀菌剂用量减少30–70%；这些数据仅用于基准参照，本研究并未测定，因为未开展比较性杀菌剂试验。上述结果为CARI的技术可操作性提供了概念验证，并明确了在计划中的多季节研究中开展正式预测评估所需的数据集和分析；不应将其解读为CARI预测准确性或减投入性能的证据。","Frontiers in Plant Physiology","2026-10-06T00:00:00Z","论文",67,{"impact":80,"substance":18,"depth":81,"authority":20,"freshness":82,"relevant":22,"comment":83},8,17,9,"方法新颖、数据翔实的木薯炭疽病早期预警概念验证研究，但仅为单点单季试点、无预测精度与减药效果验证，公共影响有限，适合作为技术前沿条目而非每日精选头条。",[85],{"name":75,"url":72},[27,28,87,88,89],"气候智慧农业","病害预警","环境传感",{"code":32,"slug":33,"name":34,"path":35,"parent_code":36},[],{"status":39,"locked":14,"version":40,"method":41,"evidence":93},{"candidates":94,"agricultural_evidence":46},[95,101,105,110],{"code":32,"score":96,"supported":46,"title":97,"abstract":98,"tags":100},3,[],[99],"decision support",[],{"code":53,"score":22,"supported":14,"title":102,"abstract":103,"tags":104},[],[],[28],{"code":106,"score":22,"supported":14,"title":107,"abstract":108,"tags":109},"H7",[],[],[87],{"code":58,"score":22,"supported":14,"title":111,"abstract":112,"tags":113},[],[],[27],[115,116],"CSIR-CRI 木薯 炭疽病","木薯炭疽病 风险指数 CARI","CSIR-CRI木薯炭疽病-4423",0,"10.3389\u002Ffphgy.2026.1898224",{"doi":119,"openalex_id":121,"authors":122,"venue":75,"cited_by_count":118,"oa_url":144,"card":145,"direction":149,"ingested_from":151},"W7220664728",[123,126,129,131,133,136,138,140,142],{"name":124,"orcid":125},"Cevdet Bulut","https:\u002F\u002Forcid.org\u002F0000-0003-1614-7990",{"name":127,"orcid":128},"Kwaku Onwona-Hwesofour Asante","https:\u002F\u002Forcid.org\u002F0000-0003-3539-3505",{"name":130,"orcid":8},"Gordon Akon-Yamga",{"name":132,"orcid":8},"Monica Ode Adu-Gyamfi",{"name":134,"orcid":135},"Zippora Appiah-Kubi","https:\u002F\u002Forcid.org\u002F0000-0001-5653-5707",{"name":137,"orcid":8},"Sarah Odame Afari",{"name":139,"orcid":8},"Josua Brett",{"name":141,"orcid":8},"Kwame Onwona Simpe",{"name":143,"orcid":8},"Reuben Nutefe Kuavedzi","https:\u002F\u002Fwww.frontiersin.org\u002Fjournals\u002Fplant-physiology\u002Farticles\u002F10.3389\u002Ffphgy.2026.1898224\u002Fpdf",{"tldr":146,"method":147,"finding":148,"direction":149,"opportunity":150},"构建木薯炭疽病阶段加权风险指数CARI，融合环境传感与目测病情，实现气候智慧型预警。","13个月微气候传感监测加15株每周病情目测，构建ERS与VSS复合指数。","CARI全程判为低风险、未发预警，仅一次发病且环境风险信号领先症状约一天。","智慧农业 \u002F 农业物联网","需多季节数据验证预警阈值与滞后关系，并开展减药对比试验评估实际投入削减效果。","openalex","2026-10-08T23:30:11.101422Z",{"id":154,"title":155,"url":156,"summary":157,"summary_zh":158,"content":8,"source_name":159,"source_url":156,"published_at":160,"category":77,"cover_url":8,"hotness":13,"is_selected":14,"score":161,"score_detail":162,"sources":166,"tags":168,"primary_topic":171,"secondary_topics":172,"classification":173,"search_phrases":207,"slug":210,"view_count":22,"doi":211,"paper":212,"created_at":224},3899,"Leveraging on Artificial Intelligence Technology in Boosting Cassava Production in Nigeria: The Gains and Constraints","https:\u002F\u002Fdoi.org\u002F10.64348\u002Fzije.2026560","Artificial Intelligence (AI) technology is reshaping agricultural landscape globally by employing data-driven, efficient, and scalable production methods. Nigeria, as the world’s largest producer of cassava, faces persistent challenges like low yield per hectare, pest infestations, climate variability, and inefficient farming techniques. This study focuses on the potential of AI technology to exploit and enhance the nation’s agricultural potentials and productivity. The study decisively examines the application of AI in cassava production in Nigeria, focusing on its gains and constraints. It adopted ways to addressing the critical challenges like climate variability, scarce access to digital farming techniques, and the need for sustainable practices. Using a descriptive and analytical approach based on secondary data from FAO, NBS, IITA, and recent scholarly publications, the study explores AI-driven technologies such as precision agriculture, machine learning, remote sensing, and automation. Findings from the study reveal that AI technology application in cassava production can significantly improve productivity, optimize resource utilization, and reduce post-harvest losses. However, adoption can be constrained by high costs, poor infrastructure, low digital literacy, and weak policy frameworks. This paper investigates how AI technology can be employed to achieve the mass production of cassava in Nigeria, while critically examining the associated gains or benefits and constraints.","人工智能（AI）技术正通过采用数据驱动、高效且可扩展的生产方式重塑全球农业格局。尼日利亚作为世界最大的木薯生产国，长期面临单位面积产量低、病虫害侵袭、气候多变以及耕作技术低效等挑战。本研究聚焦于AI技术在开发与提升该国农业潜力及生产力方面的潜力，系统考察AI在尼日利亚木薯生产中的应用，重点关注其收益与制约因素。研究探讨了应对气候多变、数字农业技术获取困难以及可持续实践需求等关键挑战的途径。基于联合国粮农组织（FAO）、尼日利亚国家统计局（NBS）、国际热带农业研究所（IITA）及近期学术文献的二手数据，采用描述性与分析性方法，探讨了精准农业、机器学习、遥感及自动化等AI驱动技术。研究发现，AI技术在木薯生产中的应用可显著提高生产力、优化资源利用并减少采后损失。然而，其推广可能受制于高成本、基础设施薄弱、数字素养偏低及政策框架不健全等因素。本文探讨如何运用AI技术实现尼日利亚木薯的大规模生产，同时批判性地审视相关收益与制约因素。","Federal University Gusau Faculty of Education Journal","2026-09-30T00:00:00Z",69,{"impact":17,"substance":163,"depth":19,"authority":164,"freshness":80,"relevant":22,"comment":165},16,12,"基于二手数据的综述性论文，系统梳理AI在尼日利亚木薯生产中的增益与制约，对非洲农业数字化有参考价值，但方法创新与一手数据有限。",[167],{"name":159,"url":156},[27,48,169,28,170],"精准农业","尼日利亚",{"code":32,"slug":33,"name":34,"path":35,"parent_code":36},[],{"status":39,"locked":14,"version":40,"method":41,"evidence":174},{"candidates":175,"agricultural_evidence":46},[176,183,190,197,203],{"code":32,"score":17,"supported":46,"title":177,"abstract":179,"tags":182},[178],"artificial intelligence",[180,181,178],"机器学习","machine learning",[48],{"code":58,"score":184,"supported":46,"title":185,"abstract":186,"tags":189},11,[],[169,187,188],"数字农业","precision agriculture",[27,169],{"code":191,"score":21,"supported":46,"title":192,"abstract":193,"tags":196},"I3",[],[194,195],"遥感","remote sensing",[],{"code":198,"score":96,"supported":46,"title":199,"abstract":200,"tags":202},"J3",[],[201],"全球农业",[],{"code":53,"score":22,"supported":14,"title":204,"abstract":205,"tags":206},[],[],[28],[208,209],"尼日利亚 木薯 人工智能","木薯 精准农业 遥感","尼日利亚木薯人工智能-3899","10.64348\u002Fzije.2026560",{"doi":211,"openalex_id":213,"authors":214,"venue":159,"cited_by_count":118,"oa_url":156,"card":217,"direction":223,"ingested_from":151},"W7214923259",[215],{"name":216,"orcid":8},"Richard Igbunu",{"tldr":218,"method":219,"finding":220,"direction":221,"opportunity":222},"分析AI在尼日利亚木薯生产中的应用收益与制约因素。","基于FAO、NBS、IITA等二手数据的描述性与分析性研究。","AI可提升木薯产量、优化资源利用并减少采后损失，但受成本、基础设施、数字素养和政策制约。","农业人工智能与决策模型","可针对小农户低成本AI工具、数字素养培训与政策框架开展实证研究。","农业遥感与作物表型","2026-10-01T23:30:29.024444Z",{"id":226,"title":227,"url":228,"summary":229,"summary_zh":230,"content":8,"source_name":231,"source_url":228,"published_at":232,"category":77,"cover_url":8,"hotness":13,"is_selected":14,"score":233,"score_detail":234,"sources":237,"tags":239,"primary_topic":242,"secondary_topics":243,"classification":244,"search_phrases":290,"slug":293,"view_count":22,"doi":294,"paper":295,"created_at":315},1980,"Green Artificial Intelligence Framework for EnergyEfficient Smart Agriculture in Tropical Environments","https:\u002F\u002Fdoi.org\u002F10.68012\u002Fair.v1i2.230","The increasing demand for sustainable agricultural production in tropical environments has encouraged the integration of environmentally responsible digital technologies to address energy consumption and resource inefficiency in smart farming systems. This study proposes a Green Artificial Intelligence framework designed to enhance energy efficiency, optimize agricultural resource utilization, and support sustainable farming practices in tropical regions characterized by high climate variability and ecological sensitivity. The research aims to develop an intelligent and adaptive agricultural model capable of reducing computational energy usage while improving environmental monitoring and decision making processes in precision agriculture. The proposed framework integrates Internet of Things sensors, machine learning algorithms, and energyaware data processing techniques to monitor soil moisture, temperature, humidity, and crop health in real time. A quantitative experimental approach was employed using simulated agricultural datasets and smart sensor data collected from tropical farming environments. The results demonstrate that the proposed Green AI framework significantly reduces energy consumption in data processing operations while maintaining high predictive accuracy and operational efficiency in smart irrigation and crop management systems. Furthermore, the framework improves environmental sustainability by minimizing excessive water and energy usage in agricultural activities. In conclusion, the study high lights the potential of Green Artificial Intelligence as an innovative solution for developing sustainable, resilient, and energyefficient smart agriculture systems that support longterm environmental conservation and food security in tropical regions.","热带环境下可持续农业生产需求的日益增长，推动了环境友好型数字技术的整合，以解决智能农业系统中的能源消耗和资源效率低下问题。本研究提出了一种绿色人工智能框架，旨在提升能源效率、优化农业资源利用，并支持气候多变性和生态敏感性显著的热带地区的可持续农业实践。研究目标是开发一种智能且自适应的农业模型，能够在减少计算能耗的同时，改善精准农业中的环境监测和决策过程。该框架整合了物联网传感器、机器学习算法和能量感知数据处理技术，实时监测土壤湿度、温度、湿度和作物健康状况。研究采用定量实验方法，利用模拟农业数据集和从热带农业环境收集的智能传感器数据。结果表明，所提出的绿色人工智能框架显著降低了数据处理操作中的能源消耗，同时保持了智能灌溉和作物管理系统中的高预测精度和运行效率。此外，该框架通过减少农业活动中的过度用水和能源使用，提升了环境可持续性。总之，本研究强调了绿色人工智能作为创新解决方案的潜力，可用于开发可持续、有韧性且节能的智能农业系统，以支持热带地区的长期环境保护和粮食安全。","OpenAlex","2026-09-09T00:00:00Z",62,{"impact":164,"substance":18,"depth":163,"authority":164,"freshness":235,"relevant":22,"comment":236},2,"提出绿色AI框架用于热带智慧农业节能，方法新颖，数据模拟，但时效性低。",[238],{"name":231,"url":228},[27,48,29,240,241],"绿色AI","节能增效",{"code":32,"slug":33,"name":34,"path":35,"parent_code":36},[],{"status":39,"locked":14,"version":40,"method":41,"evidence":245},{"candidates":246,"agricultural_evidence":46},[247,252,256,263,270,277,284],{"code":58,"score":17,"supported":46,"title":248,"abstract":250,"tags":251},[249],"smart agriculture",[169,249,188],[27],{"code":32,"score":17,"supported":46,"title":253,"abstract":254,"tags":255},[178],[180,181,178],[48],{"code":257,"score":21,"supported":46,"title":258,"abstract":259,"tags":262},"A8",[],[260,261],"粮食安全","food security",[],{"code":264,"score":21,"supported":46,"title":265,"abstract":266,"tags":269},"H2",[],[267,268],"土壤","soil",[],{"code":271,"score":21,"supported":46,"title":272,"abstract":273,"tags":276},"H3",[],[274,275],"灌溉","irrigation",[],{"code":278,"score":21,"supported":46,"title":279,"abstract":280,"tags":283},"I5",[],[281,282],"物联网","internet of things",[],{"code":285,"score":96,"supported":46,"title":286,"abstract":287,"tags":289},"H8",[],[288],"可持续农业",[],[291,292],"农业人工智能 智慧农业 热带农业 节能增效","农业人工智能 智慧农业","农业人工智能智慧农业热带农业节能增效-1980","10.68012\u002Fair.v1i2.230",{"doi":294,"openalex_id":296,"authors":297,"venue":309,"cited_by_count":118,"oa_url":228,"card":310,"direction":149,"ingested_from":151},"W7211973626",[298,301,303,306],{"name":299,"orcid":300},"Nur-Adib Maspo","https:\u002F\u002Forcid.org\u002F0000-0003-0031-1354",{"name":302,"orcid":8},"David Edmond",{"name":304,"orcid":305},"Tri Pujiati","https:\u002F\u002Forcid.org\u002F0009-0002-6067-6259",{"name":307,"orcid":308},"Meria Zakiyah Alfisuma","https:\u002F\u002Forcid.org\u002F0009-0006-6563-4495","AI Innovation and Resilience for the Environment (AIR)",{"tldr":311,"method":312,"finding":313,"direction":149,"opportunity":314},"提出绿色AI框架，提升热带智慧农业能效与资源利用。","集成物联网传感器、机器学习与能量感知数据处理技术。","显著降低数据处理能耗，保持高预测精度与运行效率。","可探索真实热带农场部署，优化模型精度与能耗平衡，或扩展至其他作物与气候区。","2026-09-09T23:30:08.343677Z",{"id":317,"title":318,"url":319,"summary":320,"summary_zh":321,"content":8,"source_name":322,"source_url":319,"published_at":323,"category":77,"cover_url":8,"hotness":13,"is_selected":14,"score":161,"score_detail":324,"sources":327,"tags":329,"primary_topic":333,"secondary_topics":339,"classification":341,"search_phrases":358,"slug":361,"view_count":118,"doi":362,"paper":363,"created_at":377},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",{"impact":80,"substance":325,"depth":17,"authority":164,"freshness":13,"relevant":22,"comment":326},21,"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。",[328],{"name":322,"url":319},[27,48,330,331,332],"小农户","植物病害识别","可解释性",{"code":334,"slug":335,"name":336,"path":337,"parent_code":338},"G2","farm-services","社会化服务与规模经营","\u002Ftopic\u002Frural-services\u002Ffarm-services","G",[340],{"code":32,"slug":33,"name":34,"path":35,"parent_code":36},{"status":39,"locked":14,"version":40,"method":41,"evidence":342},{"candidates":343,"agricultural_evidence":46},[344,349,354],{"code":334,"score":164,"supported":46,"title":345,"abstract":347,"tags":348},[346],"smallholder",[346],[330],{"code":32,"score":13,"supported":46,"title":350,"abstract":352,"tags":353},[351],"computer vision",[],[48,331],{"code":58,"score":22,"supported":14,"title":355,"abstract":356,"tags":357},[],[],[27],[359,360],"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":362,"openalex_id":364,"authors":365,"venue":322,"cited_by_count":118,"oa_url":319,"card":372,"direction":221,"ingested_from":151},"W7223137323",[366,368,370],{"name":367,"orcid":8},"Akshay Ramgoolam",{"name":369,"orcid":8},"Priya Bundhoo Appadoo",{"name":371,"orcid":8},"Jean-Noël LEspérance",{"tldr":373,"method":374,"finding":375,"direction":221,"opportunity":376},"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":379,"title":380,"url":381,"summary":382,"summary_zh":383,"content":8,"source_name":322,"source_url":381,"published_at":323,"category":77,"cover_url":8,"hotness":13,"is_selected":14,"score":161,"score_detail":384,"sources":387,"tags":389,"primary_topic":393,"secondary_topics":397,"classification":399,"search_phrases":418,"slug":421,"view_count":118,"doi":422,"paper":423,"created_at":435},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":80,"substance":385,"depth":17,"authority":164,"freshness":82,"relevant":22,"comment":386},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。",[388],{"name":322,"url":381},[27,48,390,391,392],"灌溉决策","安全保证","试验设计",{"code":271,"slug":275,"name":394,"path":395,"parent_code":396},"水资源与灌溉","\u002Ftopic\u002Fenvironment\u002Firrigation","H",[398],{"code":32,"slug":33,"name":34,"path":35,"parent_code":36},{"status":39,"locked":14,"version":40,"method":41,"evidence":400},{"candidates":401,"agricultural_evidence":46},[402,406,410,414],{"code":271,"score":19,"supported":46,"title":403,"abstract":404,"tags":405},[275],[274,275],[390],{"code":32,"score":164,"supported":46,"title":407,"abstract":408,"tags":409},[178],[99],[48],{"code":264,"score":21,"supported":46,"title":411,"abstract":412,"tags":413},[],[267,268],[],{"code":58,"score":22,"supported":14,"title":415,"abstract":416,"tags":417},[],[],[27],[419,420],"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":422,"openalex_id":424,"authors":425,"venue":322,"cited_by_count":118,"oa_url":381,"card":430,"direction":221,"ingested_from":151},"W7223131205",[426,428],{"name":427,"orcid":8},"Nassima Benkhelifa",{"name":429,"orcid":8},"Yacine Aït Ahmed",{"tldr":431,"method":432,"finding":433,"direction":221,"opportunity":434},"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":437,"title":438,"url":439,"summary":440,"summary_zh":441,"content":8,"source_name":322,"source_url":439,"published_at":323,"category":77,"cover_url":8,"hotness":13,"is_selected":14,"score":442,"score_detail":443,"sources":445,"tags":447,"primary_topic":450,"secondary_topics":451,"classification":453,"search_phrases":468,"slug":471,"view_count":118,"doi":472,"paper":473,"created_at":487},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":80,"substance":385,"depth":17,"authority":164,"freshness":13,"relevant":22,"comment":444},"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。",[446],{"name":322,"url":439},[27,48,330,448,449],"作物病害识别","模型审计",{"code":334,"slug":335,"name":336,"path":337,"parent_code":338},[452],{"code":32,"slug":33,"name":34,"path":35,"parent_code":36},{"status":39,"locked":14,"version":40,"method":41,"evidence":454},{"candidates":455,"agricultural_evidence":46},[456,460,464],{"code":334,"score":19,"supported":46,"title":457,"abstract":458,"tags":459},[346],[330,346],[330],{"code":32,"score":13,"supported":46,"title":461,"abstract":462,"tags":463},[351],[],[448,48],{"code":58,"score":22,"supported":14,"title":465,"abstract":466,"tags":467},[],[],[27],[469,470],"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":472,"openalex_id":474,"authors":475,"venue":322,"cited_by_count":118,"oa_url":439,"card":482,"direction":221,"ingested_from":151},"W7223088030",[476,478,480],{"name":477,"orcid":8},"Abdi-Basid Idriss Farah",{"name":479,"orcid":8},"Amina Abdillahi Guedi",{"name":481,"orcid":8},"Moussa Omar Youssouf",{"tldr":483,"method":484,"finding":485,"direction":221,"opportunity":486},"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"]