[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-123269-en":3,"doc-seo-123269-105":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":4,"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},123269,1099514067438,"River Wang","https://ap-avatar.wpscdn.com/avatar/100002539ee87300030?x-image-process=image/resize,m_fixed,w_180,h_180&k=1780474512215547542",8,"Research & Report","Machine Learning for Gap-Filling of Missing Data in Time Series - A Study of Sea Surface Temperature in Trondheimsfjorden","Incomplete datasets hinder ocean research and restrict understanding of key ocean variables. This thesis applies machine learning to fill gaps in time-series, using sea surface temperature measurements collected by the Munkholmen OceanLab buoy in Trondheimsfjorden. The work gathers and prepares multiple environmental datasets and designs three gap-filling methods: linear interpolation, ARIMA, and recurrent neural networks. The methods are evaluated and compared, showing distinct strengths and weaknesses, and improved neural performance when multiple related variables are used together.","Masteroppgave  \nNT NU  \nNorges te kn isk-naturvitenskapelige universitet Fakultet for informasjonstekno log i og elektroteknikk Institutt for datatekno log i og informatikk  \nHans Eivind Johnsen Skinstad  \nMachine Learning for Gap-Filling of Missing Data in Time Series  \nA Study of Sea Surface Temperature in Trondheimsfjorden  \nMasteroppgave i Master i datateknologi Veileder: Ole Jakob Mengshoel  \nMedveileder: Ute Brönner, Lara Veylit Juni 2024  \nHans Eivind Johnsen Skinstad  \nMachine Learning for Gap-Filling of Missing Data in Time Series  \nA Study of Sea Surface Temperature in Trondheimsfjorden  \nMaster’s thesis in Master i datateknologi Supervisor: Ole Jakob Mengshoel  \nCo-supervisor: Ute Brönner, Lara Veylit June 2024  \nNorwegian University of Science and Technology  \nAbstract  \nIncomplete datasets create difficulties for ocean research and can hamper our ability to understand the role of important ocean variables. This thesis explores the application of using machine learning to fill gaps in time-series with a focus on thesea surface temperature measurements captured by the Munkholmen OceanLab buoy in Trondheimsfjorden. Our work includes identifying, gathering and preparing relevant datasets of a number of environmental variables and designing three different methods to perform gap-filling of missing data. These methods based on linear interpolation, the statistical ARIMA method, and recurrent neural networks, are evaluated and their performances compared. The results of this work show that each method has its unique strengths and weaknesses. In addition, we observe how the performance of our neural network model improves when utilizing multiple, related variables simultaneously. The project demonstrates that machine learning methods are a viable option for to filling in missing data while highlighting the importance of complete datasets for environmental research.  \nPreface  \nThis project was conducted as part of the course TDT4900-Computer Science, Master’s Thesis at the Norwegian University of Science and Technology (NTNU) . This thesis was written and carried out by Hans Eivind Johnsen Skinstad, master student at the Department of Computer Science, Faculty of Information Technology and Electrical Engineering.  \nI would like to give a thanks to my supervisor Ole Jakob Mengshoel for helping out when needed and providing valuable advice and feedback.  \nI also want to thank my co-supervisors Ute Brönner and Lara Veylit at SINTEF Ocean for their contributions to this project: their guidance and availability, and for facilitating my many visits to SINTEF Ocean.  \nAnother thanks to Morten Omholt Alver at NTNU for sharing his knowledge on ocean dynamics and providing support on simulated data, and a last thanks to Raymond Nepstad at SINTEF Ocean for his valuable feedback and support during this project.  \nParts of this thesis are taken from or based on my submitted project assignment in the course TDT4501 - Computer Science, Specialization Project with the title\"Machine Learning for Effective Ocean Data Analysis \" [1] . Section 2.1 is entirely a copy of major parts of section 2 . 1 in the project assignment. The same goes for section 2.3 being a copy of a major part of section 2.4 in the project assignment, but with a few minor changes. Parts of section 3 .2 is a copy of section 3 . 1 in the project assignment, with the rest being rewritten or built upon it. Parts of the very first paragraph of section 5.2.2 was based on section 2.4.1 in the project assignment.  \nContents  \nAbstract i  \nPreface ii  \nContents v  \nList of Figures v  \nAbbreviations viii  \n1 Introduction 1  \n1.1 Motivation ................................ 1  \n1.2 Partner ................................. 1  \n1.3 Research goal .............................. 2  \n1.4 Report Structure ............................ 3  \n2 Background and theory 5  \n2.1 Oceanography .............................. 5  \n2.1.1 Ocean dynamics ......................... 5  \n2.2 Time-series .....","cbCaikCzIYtRMmih","https://ap.wps.com/l/cbCaikCzIYtRMmih","pdf",9763401,1,83,"English","en",105,"# Abstract\n# Preface\n# Contents\n# List of Figures\n# Abbreviations\n# 1 Introduction\n## 1.1 Motivation\n## 1.2 Partner\n## 1.3 Research goal\n## 1.4 Report Structure\n# 2 Background and theory\n## 2.1 Oceanography\n## 2.2 Time-series\n## 2.3 Artificial neural networks\n## 2.4 Types of ANNs\n## 2.5 Time-series gap-filling\n# 3 Methods\n## 3.1 Dataset\n## 3.2 Selecting a machine learning library\n## 3.3 Gap-filling methods\n## 3.4 Measuring the performance of the models\n## 3.5 Deciding hyperparameters\n# 4 Results","[{\"question\":\"What problem does the thesis address in ocean research?\",\"answer\":\"It addresses how incomplete datasets create difficulties for ocean research and can limit understanding of important ocean variables.\"},{\"question\":\"Which sea surface temperature data source is used?\",\"answer\":\"The study uses sea surface temperature measurements captured by the Munkholmen OceanLab buoy in Trondheimsfjorden.\"},{\"question\":\"How are missing-data gap-filling methods evaluated and compared?\",\"answer\":\"Three methods are implemented—linear interpolation, ARIMA, and recurrent neural networks—and their performance is compared using metrics such as mean squared and absolute error and Pearson correlation.\"}]","Machine Learning for Gap-Filling of Missing Data in Time Series - A Study of Sea Surface Temperature in Trondheimsfjorden | PDF",1785815603,209,{"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},"machine-learning-for-gap-filling-of-missing-data-in-time-series-a-study-of-sea-surface-temperature-in-trondheimsfjorden","",{"@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/machine-learning-for-gap-filling-of-missing-data-in-time-series-a-study-of-sea-surface-temperature-in-trondheimsfjorden/123269/",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":4},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"What problem does the thesis address in ocean research?","Question",{"text":75,"@type":76},"It addresses how incomplete datasets create difficulties for ocean research and can limit understanding of important ocean variables.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"Which sea surface temperature data source is used?",{"text":80,"@type":76},"The study uses sea surface temperature measurements captured by the Munkholmen OceanLab buoy in Trondheimsfjorden.",{"name":82,"@type":73,"acceptedAnswer":83},"How are missing-data gap-filling methods evaluated and compared?",{"text":84,"@type":76},"Three methods are implemented—linear interpolation, ARIMA, and recurrent neural networks—and their performance is compared using metrics such as mean squared and absolute error and Pearson correlation.","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"]