Agriculture has become an information-intensive industry. In the production of crops and animals, precision agriculture approaches have resulted in the collection of spatially and temporally dense datasets by farmers and agricultural researchers. These big datasets, often characterized by extensive nonlinearities and interactions, are often best analyzed using machine learning (ML) or other artificial intelligence (AI) approaches. In this article, we review several case studies where ML has been used to model aspects of agricultural production systems and provide information useful for farm-level management decisions. These studies include modeling animal feeding behavior as a predictor of stress or disease, providing information important for developing precise and efficient irrigation systems, and enhancing tools used to recommend optimum levels of nitrogen fertilization for corn. Taken together, these examples represent the current abilities and future potential for AI applications in agricultural production systems.

中文翻译：

---

农场里的人工智能

农业已成为信息密集型产业。在作物和动物的生产中，精准农业方法导致农民和农业研究人员收集了空间和时间密集的数据集。这些大数据集通常以广泛的非线性和交互为特征，通常最好使用机器学习 (ML) 或其他人工智能 (AI) 方法进行分析。在本文中，我们回顾了几个案例研究，其中 ML 已被用于对农业生产系统的各个方面进行建模，并提供对农场级管理决策有用的信息。这些研究包括将动物喂养行为建模为压力或疾病的预测指标，为开发精确有效的灌溉系统提供重要信息，以及用于推荐玉米最佳氮肥水平的增强工具。总之，这些例子代表了人工智能在农业生产系统中应用的当前能力和未来潜力。