[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-122087-en":3,"doc-seo-122087-105":30,"detail-sidebar-cat-0-en-105":91},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":20,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":27,"seo_description":14,"update_tm":28,"read_time":29},122087,8796095461610,"Oliver","https://ap-avatar.wpscdn.com/davatar_276721f389ce27ea32af1340a28f341c",8,"Research & Report","Soil Science-Informed Machine Learning - Research overview and methodologies","Machine learning (ML) applications in soil science have expanded rapidly over the past two decades, driven by data-driven research to improve soil security. Most work emphasizes predicting soil properties, especially soil organic carbon, and increasing the accuracy of digital soil mapping. Key limitations remain data scarcity and limited interpretability. The paper advocates Soil Science-Informed ML (SoilML) models that embed soil science knowledge into training to improve reliability and generalisation.","UC Merced  \nUC Merced Previously Published Works  \nTitle  \nSoil Science-Informed Machine Learning  \nPermalink  \n[https://escholarship.org/uc/item/8qf2n8zs](https://escholarship.org/uc/item/8qf2n8zs)  \nAuthors  \nMinasny, Budiman Bandai, Toshiyuki Ghezzehei, Teamrat A et al.  \nPublication Date  \n2024-12-01  \nDOI  \n10.1016/j.geoderma.2024.117094  \nPeer reviewed  \n[eScholarship.org](eScholarship.org) Powered by the California Digital Library  \nUniversity of California  \n1 Soil Science-Informed Machine Learning  \n2  \n3 Budiman Minasny 1 , Toshiyuki Bandai2 , Teamrat A. Ghezzehei3 ; Yin-Chung Huang 1 , 4 Yuxin Ma4 , Alex. B. McBratney 1 , Wartini Ng 1 , Sarem Norouzi5 , Jose Padarian 1 ,  \n5 Rudiyanto6 , Amin Sharififar 1 , Quentin Styc 1 , Marliana Widyastuti 1 6  \n7  \n8  \n9 1 School of Life and Environmental Sciences, The University of Sydney, NSW 2006, 10 Australia.  \n11 2 Earth and Environmental Sciences Area, Lawrence Berkeley National Laboratory, 12 Berkeley, CA 94720, USA.  \n13 3 Life & Environmental Sciences Department, University of California, Merced, CA  \n14 95343, USA.  \n15 4 New South Wales Department of Climate Change, Energy, the Environment and  \n16 Water, Parramatta, NSW 2150, Australia.  \n17 5 Department of Agroecology, Aarhus University, 8830 Tjele, Denmark.  \n18 6 Faculty of Fisheries and Food Science, Universiti Malaysia Terengganu, 21030 Kuala  \n19 Nerus, Terengganu, Malaysia.  \n20  \n21 Abstract  \n22 Machine learning (ML) applications in soil science have significantly increased over  \n23 the past two decades, reflecting a growing trend towards data-driven research  \n24 addressing soil security. This extensive application has mainly focused on enhancing  \n25 predictions of soil properties, particularly soil organic carbon, and improving the  \n26 accuracy of digital soil mapping (DSM) . Despite these advancements, the  \n27 application of ML in soil science faces challenges related to data scarcity and the  \n28 interpretability of ML models. There is a need for a shift towards Soil Science-  \n29 Informed ML (SoilML) models that use the power of ML but also incorporate soil  \n30 science knowledge in the training process to make predictions more reliable and  \n31 generalisable. This paper proposes methodologies for embedding ML models with  \n32 soil science knowledge to overcome current limitations. Incorporating soil science  \n33 knowledge into ML models involves using observational priors to enhance training  \n34 datasets, designing model structures which reflect soil science principles, and  \n35 supervising model training with soil science-informed loss functions. The informed  \n36 loss functions include observational constraints, coherency rules such as  \n37 regularisation to avoid overfitting, and prior or soil-knowledge constraints that  \n38 incorporate existing information about the parameters or outputs. By way of  \n39 illustration, we present examples from four fields: digital soil mapping, soil  \n40 spectroscopy, pedotransfer functions, and dynamic soil property models. We discuss  \n41 the potential to integrate process-based models for improved prediction, the use of  \n42 physics-informed neural networks, limitations, and the issue of overparametrisation.  \n43 These approaches improve the relevance of ML predictions in soil science and  \n44 enhance the models' ability to generalise across different scenarios while  \n45 maintaining soil science principles, transparency and reliability.  \n46 1. Introduction  \n47 The 2024 Nobel Prize in Physics was awarded to researchers who utilised physics- 48 based tools to develop methods that advance machine learning through artificial  \n49 neural networks. Over the past two decades, the use of machine learning (ML) in  \n50 soil research has surged. In 2023, an average of 8 papers per day were published on  \n51 topics related to “machine learning” and “soil”(Scopus, June 2024) . These  \n52 advancements, highlight the growing importance of ML in various scientific field","cbCaioYgLjBntWyV","https://ap.wps.com/l/cbCaioYgLjBntWyV","pdf",2614565,1,66,"English","en",105,"# Abstract\n# Introduction\n## Motivations and growth of ML in soil research\n## Limitations of conventional ML approaches","[{\"question\":\"What is the main issue with using conventional machine learning in soil science?\",\"answer\":\"Conventional ML often struggles with limited data availability and with interpretability, and learned patterns may not generalise to unseen scenarios.\"},{\"question\":\"What does Soil Science-Informed ML (SoilML) aim to achieve?\",\"answer\":\"SoilML incorporates soil science knowledge into model training so predictions become more reliable and generalisable while maintaining soil-science principles.\"},{\"question\":\"How are soil science constraints integrated into ML models in this work?\",\"answer\":\"By using observational priors to strengthen training data, designing model structures that reflect soil science principles, and supervising training with soil science-informed loss functions.\"}]","Soil Science-Informed Machine Learning - Research overview and methodologies | PDF",1785808753,166,{"code":4,"msg":31,"data":32},"ok",{"site_id":24,"language":23,"slug":33,"title":13,"keywords":34,"description":14,"schema_data":35,"social_meta":86,"head_meta":88,"extra_data":90,"updated_unix":28},"soil-science-informed-machine-learning-research-overview-and-methodologies","",{"@graph":36,"@context":85},[37,54,68],{"@type":38,"itemListElement":39},"BreadcrumbList",[40,44,48,51],{"item":41,"name":42,"@type":43,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":45,"name":46,"@type":43,"position":47},"https://docshare.wps.com/document/","Document",2,{"item":49,"name":12,"@type":43,"position":50},"https://docshare.wps.com/document/research-report/",3,{"item":52,"name":13,"@type":43,"position":53},"https://docshare.wps.com/document/soil-science-informed-machine-learning-research-overview-and-methodologies/122087/",4,{"url":52,"name":13,"@type":55,"author":56,"headline":13,"publisher":58,"fileFormat":61,"inLanguage":23,"description":14,"dateModified":62,"datePublished":62,"encodingFormat":61,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":57},"Person",{"url":41,"name":59,"@type":60},"DocShare","Organization","application/pdf","2026-08-04",true,{"@type":65,"interactionType":66,"userInteractionCount":20},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"What is the main issue with using conventional machine learning in soil science?","Question",{"text":75,"@type":76},"Conventional ML often struggles with limited data availability and with interpretability, and learned patterns may not generalise to unseen scenarios.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"What does Soil Science-Informed ML (SoilML) aim to achieve?",{"text":80,"@type":76},"SoilML incorporates soil science knowledge into model training so predictions become more reliable and generalisable while maintaining soil-science principles.",{"name":82,"@type":73,"acceptedAnswer":83},"How are soil science constraints integrated into ML models in this work?",{"text":84,"@type":76},"By using observational priors to strengthen training data, designing model structures that reflect soil science principles, and supervising training with soil science-informed loss functions.","https://schema.org",{"og:url":52,"og:type":87,"og:title":13,"og:site_name":59,"og:description":14},"article",{"robots":89,"canonical":52},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":92},[93,97,101,105,110,115,120,123,128,131,135],{"id":20,"doc_module":4,"doc_module_name":46,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":53,"doc_module":4,"doc_module_name":46,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":46,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":111,"doc_module":4,"doc_module_name":46,"category_name":112,"show_sort_weight":113,"slug":114},6,"Technology",50,"technology",{"id":116,"doc_module":4,"doc_module_name":46,"category_name":117,"show_sort_weight":118,"slug":119},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":121,"slug":122},30,"research-report",{"id":124,"doc_module":4,"doc_module_name":46,"category_name":125,"show_sort_weight":126,"slug":127},9,"Religion & Spirituality",20,"religion-spirituality",{"id":126,"doc_module":4,"doc_module_name":46,"category_name":129,"show_sort_weight":126,"slug":130},"World Cup","world-cup",{"id":132,"doc_module":4,"doc_module_name":46,"category_name":133,"show_sort_weight":132,"slug":134},10,"Lifestyle","lifestyle",{"id":136,"doc_module":4,"doc_module_name":46,"category_name":137,"show_sort_weight":106,"slug":138},19,"General","general"]