arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2306.09302cs.LGcs.AI
收藏

土壤科学中知识引导的表示学习与因果结构学习

Knowledge Guided Representation Learning and Causal Structure Learning in Soil Science

University of Minnesota(明尼苏达大学);Microsoft Research(微软研究院);University of Glasgow(格拉斯哥大学);Rothamsted Research(洛桑研究所)

Somya Sharma, Swati Sharma, Licheng Liu, Rishabh Tushir, Andy Neal, Robert Ness, John Crawford, Emre Kiciman, Ranveer Chandra

AI总结:

针对土壤科学中观测数据成本高、模拟数据真实性不足的问题,提出KGRCL框架,结合两类数据优化表示学习与因果结构学习,提升土壤碳影响预测的性能。

摘要:

对土壤的深入理解可推动更可持续的土地利用实践。然而,由于不同土壤过程间存在复杂的相互作用限制了我们对土壤的认知,土壤被称为复杂的活性介质。基于过程的模型与观测数据分析是提升土壤过程认知的两大途径:采集观测数据成本高昂,但能反映真实世界的规律;而基于过程的模型可生成充足的合成数据,却可能无法代表实际情况。我们提出知识引导的表示学习与因果结构学习框架(KGRCL),以加速土壤科学领域的科学发现。该框架通过与观测土壤过程进行条件分布匹配,优化了模拟土壤过程的表示学习;同时利用观测与模拟两类数据,学习土壤过程间的因果结构。与其他因果发现方法生成的因果图相比,该方法学习到的因果图更符合真实情况。此外,学习到的因果图被用于监督学习场景,以预测施肥和气候变化对土壤碳的影响。我们在5个不同地点开展实验,结果表明该方法在样本外和少样本设置下的预测性能均有所提升。

英文摘要:

An improved understanding of soil can enable more sustainable land-use practices. Nevertheless, soil is called a complex, living medium due to the complex interaction of different soil processes that limit our understanding of soil. Process-based models and analyzing observed data provide two avenues for improving our understanding of soil processes. Collecting observed data is cost-prohibitive but reflects real-world behavior, while process-based models can be used to generate ample synthetic data which may not be representative of reality. We propose a framework, knowledge-guided representation learning, and causal structure learning (KGRCL), to accelerate scientific discoveries in soil science. The framework improves representation learning for simulated soil processes via conditional distribution matching with observed soil processes. Simultaneously, the framework leverages both observed and simulated data to learn a causal structure among the soil processes. The learned causal graph is more representative of ground truth than other graphs generated from other causal discovery methods. Furthermore, the learned causal graph is leveraged in a supervised learning setup to predict the impact of fertilizer use and changing weather on soil carbon. We present the results in five different locations to show the improvement in the prediction performance in out-of-sample and few-shots setting.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑