[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-120952-en":3,"doc-seo-120952-105":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},120952,549758252649,"Ivy","https://ap-avatar.wpscdn.com/avatar/8000253669c5317157?_k=1778319167496531819",8,"Research & Report","Knowledge-Guided Machine Learning for Spatiotemporal Environmental Data Analysis - Dissertation","Ubiquitous environmental monitoring has made large-scale spatiotemporal data available, but extracting reliable information for sustainability tasks remains challenging. This dissertation applies machine learning with domain-knowledge guidance to analyze spatiotemporal environmental data. It introduces novel methods for soil moisture gap-filling and crop yield prediction, including a two-layer model inspired by radar-radiometer fusion retrieval and a semi-supervised self-attentive approach using global spatiotemporal representations and sampled contexts. Experiments show improved accuracy and robustness under training data scarcity.","KNOWLEDGE-GUIDED MACHINE LEARNING FOR SPATIOTEMPORAL ENVIRONMENTAL DATA ANALYSIS  \nA Dissertation  \nby  \nHANZI MAO  \nSubmitted to the Office of Graduate and Professional Studies of Texas A&M University  \nin partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY  \nChair of Committee, Nicholas Duffield Committee Members, Binayak Mohanty  \nDilma Da Silva  \nXia Hu  \nHead of Department, Scott Shaefer  \nAugust 2020  \nMajor Subject: Computer Science  \nCopyright 2020 Hanzi Mao  \nABSTRACT  \nWith the emergence of ubiquitous environmental monitoring systems in the past few decades, we are gaining unprecedented ability to collect vast amounts of spatiotemporal environmental data. However, it still remains a challenge to mine information from the spatiotemporal data for many environmental sustainability tasks. This dissertation focuses on utilizing machine learning to analyze spatiotemporal environmental data with the guidance of domain knowledge and specifically proposes several novel algorithms for the soil moisture gap-filling task and crop yield prediction task.  \nFirst, we study the problem of soil moisture gap-filling. Large spatiotemporal gaps can be incurred for daily soil moisture product that adopts the radar-radiometer fusion approach. This is normally due to the relatively low revisit schedule and the associated poor spatiotemporal coverage of radar observations. Gap-fill high resolution soil moisture in regional scale for remote sensing soil moisture product is however a great challenge. It requires models learned at neighboring regions to produce predictions at anew region with reasonable accuracy. To address this issue, we propose a novel two-layer machine learning-based algorithm that is motivated by the soil moisture radar-radiometer fusion retrieval algorithm. It predicts the brightness temperature and subsequently the soil moisture at gap areas. Compared with the traditional one-layer machine learning approach, this two-layer approach shows superior performance in extensive experiments at four study areas with distinct climate regimes.  \nWe then focus on information mining from the multi-channel geo-spatiotemporal data. Existing approaches adopt various dimensionality reduction techniques without fully taking advantage of the data. In addition, the lack of labeled training data raises another challenge for modeling such data. We propose a novel semi-supervised self-attentive model  \nthat learns global spatiotemporal representations for prediction tasks. Spatial and temporal variations in the geo-spatiotemporal data are extracted to produce accurate predictions. To overcome the data scarcity issue, we introduce sampled spatial and temporal context that naturally reside in the largely-available unlabeled geo-spatiotemporal data. The proposed algorithm is validated specifically on a large-scale real-world crop yield prediction task. The results show that our semi-supervised self-attentive model outperforms existing state-of-the-art yield prediction methods and its counterpart, the supervised-only self-attentive model, especially under the stress of training data scarcity.  \nDEDICATION  \nTo my parents, Bintao Mao and Chengnian Xia.  \nACKNOWLEDGMENTS  \nI would like to devote my deepest gratitude to my advisor Dr. Nicholas Duffield.  \nDr. Duffield took me as his PhD student at my hardest time and has been providing full support to my research work since then. He was the person who introduced me to my research area, applying machine learning in geo-spatiotemporal data analysis, which I have been passionate about and truly believe its potential impact in the future. Starting research work in this area, however, was not easy for me. Not only because I had litter experience in machine learning back in 2016, but also it required extensive work to understand the problems and challenges in this interdisciplinary area itself. Dr. Duffield offered tremendous encouragement and support during those times. He continuously fue","cbCains6VLVnxprn","https://ap.wps.com/l/cbCains6VLVnxprn","pdf",31042963,1,129,"English","en",105,"# Abstract\n## Soil moisture gap-filling\n## Semi-supervised self-attentive modeling for crop yield prediction\n## Contributions and experimental validation","[{\"question\":\"What core problem does the dissertation address?\",\"answer\":\"It addresses mining useful information from spatiotemporal environmental data for sustainability tasks, focusing on soil moisture gap-filling and crop yield prediction.\"},{\"question\":\"How does the proposed soil moisture gap-filling approach work?\",\"answer\":\"It uses a novel two-layer machine learning algorithm motivated by radar-radiometer fusion retrieval, first predicting brightness temperature and then soil moisture in gap areas.\"},{\"question\":\"What modeling strategy is used for crop yield prediction and how is data scarcity handled?\",\"answer\":\"It proposes a semi-supervised self-attentive model to learn global spatiotemporal representations, leveraging sampled spatial and temporal context from largely available unlabeled data to mitigate limited labeled training samples.\"}]","Knowledge-Guided Machine Learning for Spatiotemporal Environmental Data Analysis - Dissertation | PDF",1785733011,325,{"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},"knowledge-guided-machine-learning-for-spatiotemporal-environmental-data-analysis-dissertation","",{"@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/knowledge-guided-machine-learning-for-spatiotemporal-environmental-data-analysis-dissertation/120952/",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","2026-08-03",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 core problem does the dissertation address?","Question",{"text":75,"@type":76},"It addresses mining useful information from spatiotemporal environmental data for sustainability tasks, focusing on soil moisture gap-filling and crop yield prediction.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"How does the proposed soil moisture gap-filling approach work?",{"text":80,"@type":76},"It uses a novel two-layer machine learning algorithm motivated by radar-radiometer fusion retrieval, first predicting brightness temperature and then soil moisture in gap areas.",{"name":82,"@type":73,"acceptedAnswer":83},"What modeling strategy is used for crop yield prediction and how is data scarcity handled?",{"text":84,"@type":76},"It proposes a semi-supervised self-attentive model to learn global spatiotemporal representations, leveraging sampled spatial and temporal context from largely available unlabeled data to mitigate limited labeled training samples.","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"]