Conference: Frontiers in artificial intelligence and machine learning for ecology: 17-20 October 2022

生态学和人工智能理论都是关于理解复杂系统如何工作，并定义系统的可预测性。该奖项将支持一个研讨会，召集数据科学、人工智能、计算机科学和生态等领域的专家。其目的是确定一致的研究前沿，提炼通用词汇，并支持交叉受精支持人工智能(AI)和机器学习(ML)的发展，以应对利用生态学理论和数据同化产生对复杂系统属性(如生态系统弹性和病原体出现)的预测性理解方面的邪恶挑战。人工智能和机器学习已被誉为利用大数据和生成复杂系统的假设和预测的革命性方法。生态学中的传统理论和以过程为基础的方法可以成为学习的强大工具，但往往局限于低维分析。然而，生态系统在空间、时间和相互作用的网络中都是复杂和多维的-ML和AI发展的当前方向正在探索理解这种复杂系统中的可预测性的方法。尽管机器学习方法在生态系统中描述模式、过程和预测的新兴使用，但用于批判性评估这些方法如何与生态理论发展和过程理解相一致的资源稀缺。同样，生态学家与开发AI和ML的计算机和数据科学家互动的机会也很有限，这些互动是交叉受精的关键机会，可以探索当前实践之外的东西，并完善共同的目标。本次研讨会将以广为人知的形式为两个社区提供两个已出版的交付成果，将突出生态、生态系统、计算机和数据科学受众在人工智能和ML发展方面的融合机会。该奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

Ecological and artificial intelligence theories are both about understanding how complex systems work and defining system predictability. This award will support a workshop convening experts at the nexus of data science, artificial intelligence, computer science, and ecology. The aim is to identify aligned research frontiers, refine a common vocabulary, and support cross-fertilization supporting artificial intelligence (AI) and machine learning (ML) development to target wicked challenges in leveraging ecological theory and data assimilation to generate predictive understanding of complex system properties such as ecosystem resilience and pathogen emergence Artificial intelligence and machine learning have been heralded as revolutionary approaches for harnessing big data and generating hypotheses and predictions for complex systems. Traditional theory and process-based approaches in ecology can be powerful tools for learning but are often limited to low-dimensional analyses. Yet ecological systems are complex and multi-dimensional in space, in time, and in networks of interactions - and current directions in ML and AI development are exploring approaches to understand predictability in such complex systems.Despite burgeoning use of machine learning approaches to describe pattern, process, and predictions in ecological systems, there are a paucity of resources for critically evaluating how these approaches can align with ecological theory development and process understanding. Likewise, there are limited opportunities for ecologists to interact with the computer and data scientists developing AI and ML, and these interactions are a critical opportunity for cross-fertilization to explore beyond current practices and refine shared targets. This workshop will produce two published deliverables in widely known forms for both communities that will highlight convergent opportunities for AI and ML development for ecological, ecosystems, computer, and data science audiences.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.