[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-123660-en":3,"doc-seo-123660-105":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},123660,962075006959,"Anda","https://ap-avatar.wpscdn.com/avatar/e0002397efbe92a78e?_k=1776741047341049297",8,"Research & Report","Machine Learning Methods for Autonomous Classification and Decision Making - Thesis","This PhD thesis develops machine learning methods for autonomous classification and decision making through two focused case studies: traffic speed prediction and cancer bone lesion segmentation. For traffic speed prediction, an efficient capsule network replaces pooling with dynamic routing to preserve local information and improves over state-of-the-art CNNs. Because CNN/CapsNet uncertainties are not directly analyzed, two Gaussian process frameworks are introduced to quantify uncertainty via confidence intervals and noise studies, including a spatio-temporal deep kernel CNN-GP. For bone segmentation, data generation and a GAN-based reconstruction enable lesion localization by image subtraction.","University of She􀀎eld  \nMachine Learning Methods for Autonomous Classi􀀌cation and Decision Making  \nYifei Zhu  \n1st Supervisor: Lyudmila Mihaylova 2nd Supervisor: Michael Balikhin  \nA thesis submitted in partial ful􀀌lment of the requirements for the degree of Doctor of Philosophy  \nDepartment of Automatic Control and Systems Engineering  \ni  \nDeclaration  \nAll sentences or passages quoted in this document from other people's work have been speci􀀌cally acknowledged by clear cross-referencing to author, work and page(s) . Any illustrations that are not the work of the author of this report have been used with the explicit permission of the originator and are speci􀀌cally acknowledged. I understand that failure to do this amounts to plagiarism and will be considered grounds for failure.  \nName:  \n\n| Signature: |\n| --- |\n| Date: |\n\nAcknowledgements  \nCompleting the PhD study over the last four years was a challenging journey. However, I would like to thank everyone who encouraged me and provided help through the journey.  \nFirst, I would like to thank my supervisor, Prof. L. Mihaylova, for providing guidance, feedback and encouragements. Second I would like to thank Dr. Peng Wang and Dr. Xingchi Liu who are co-authors of some of my publications and the journal paper under preparation, and who provided great help and guidance to the project conducted the thesis.  \nSecondly, I would like to express my heartfelt appreciation to my family. Thank my parents for fully supporting my further study as a PhD student. Thank my wife for making those sacri􀀌ces and accompanying me through those years.  \nAbstract  \nThis thesis focuses on developing machine learning methods for autonomous classi􀀌 -cation and decision making, especially on two case studies: tra􀀎c speed prediction and cancer bone segmentation. For tra􀀎c speed prediction, the convolutional neural network (CNN) achieves state-of-the-art results in complex tra􀀎c networks. However, the pooling layers cause the loss of information within the data. This thesis proposes an e􀀎cient capsule network for tra􀀎c speed prediction. The proposed capsule network replaces the pooling layer with capsules connected by dynamic routing and encodes the features and probability of those features showing on the local region. The proposed capsule network provides outperformed results compared to state-of-the-art CNNs. However, the CNN and capsule network (CapsNet) are parametric models and the uncertainty is, thus, not analysed. Two Gaussian process (GP) frameworks are proposed for tra􀀎c speed prediction, equipping the CNN with the ability to quantify uncertainty. The 􀀌rst framework proposes to equate a state-of-the-art CNN with a shallow GP. The proposed approach is evaluated and the uncertainty is analysed by applying the con􀀌dence interval. In addition, the impact of the noise is investigated by adding a di􀀋erent level of noise. The second framework is a novel deep kernel CNN-GP framework with spatio-temporal kernels, allowing it to abstract high-level features and consider both time and space. The proposed CNN-GP framework is validated and evaluated using CO2 concentration and tra􀀎c prediction for the short-term and long-term. An e􀀎cient uniform error bound is proposed and evaluated with simulated and real data. For cancer bone segmentation, machine learning methods are proposed to seg-  \niv  \nment bone lesions in cancer-induced bone disease from Micro Computed Tomography (µCT) images, which brings a new perspective of dealing with bone caner segmentation. The performances are evaluated and their e􀀋ectiveness is compared. Due to the limited number of datasets and the lack of labelled lesions within the dataset, an approach to generate simulated data is proposed. With an enhanced dataset, a generative adversarial network is proposed to reconstruct the bone with a lesion to a healthy bone. Consequently, the location of the lesion can be obtained by subtracting the original image from the reconstructed imag","cbCaiuPxy5loec2T","https://ap.wps.com/l/cbCaiuPxy5loec2T","pdf",3883005,1,184,"English","en",105,"# Contents\n## List of Symbols\n## List of Abbreviations\n## List of Figures\n## List of Tables\n## 1 Introduction\n## 1.1 Objective\n## 1.2 Thesis Outline\n## 1.3 Key Contributions\n## 1.4 Publications\n## 2 Literature Review\n## 2.1 Backgrounds\n## 2.2 Methods for Traffic Prediction\n## 2.2.1 Segment-based Regression Kriging (SRK)\n## 2.2.2 Auto Regressive Integrated Moving Average (ARIMA)\n## 2.2.3 Bayesian Combined Neural Network (BCNN)\n## 2.2.4 Long Short-term Memory Neural Network (LSTM-NN)\n## 2.2.5 Convolutional Neural Network for Traffic Prediction\n## 2.2.6 Structural Recurrent Neural Network for Traffic Prediction\n## 2.2.7 Support Vector Machines\n## 2.2.8 Gaussian Process Methodology\n## 2.3 Introduction of Uncertainty Quantification\n## 2.3.1 Confidence Interval\n## 2.3.2 Variational Lower Bound\n## 2.4 Fast Edge Detection Using Structured Forests for Classification and Segmentation\n## 2.4.1 Random Decision Forests\n## 2.4.2 Structured Random Forests\n## 2.4.3 Edge Detection\n## 2.5 Summary\n## 3 Machine Learning for Traffic Prediction\n## 3.1 Introduction\n## 3.2 Machine Learning Frameworks for Short-term Traffic Prediction\n## 3.2.1 Capsule Network\n## 3.2.2 Deep Convolutional Neural Network as Shallow Gaussian Process\n## 3.2.3 Performance Evaluation\n## 3.2.4 Conclusion","[{\"question\":\"What problem does this thesis address for autonomous decision making?\",\"answer\":\"It focuses on developing machine learning methods for autonomous classification and decision making, demonstrated through traffic speed prediction and cancer bone lesion segmentation tasks.\"},{\"question\":\"How is uncertainty handled for traffic speed prediction models?\",\"answer\":\"Two Gaussian process frameworks are proposed to quantify uncertainty, including a shallow GP equivalence of a CNN and a deep kernel CNN-GP with spatio-temporal kernels, evaluated using confidence intervals and noise analysis.\"},{\"question\":\"How are cancer bone lesions localized in the proposed segmentation approach?\",\"answer\":\"An enhanced dataset is generated and a generative adversarial network reconstructs healthy and lesion-affected bones; lesion location is obtained by subtracting the original image from the reconstructed image.\"}]","Machine Learning Methods for Autonomous Classification and Decision Making - Thesis | PDF",1785817888,464,{"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-methods-for-autonomous-classification-and-decision-making-thesis","",{"@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-methods-for-autonomous-classification-and-decision-making-thesis/123660/",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 this thesis address for autonomous decision making?","Question",{"text":75,"@type":76},"It focuses on developing machine learning methods for autonomous classification and decision making, demonstrated through traffic speed prediction and cancer bone lesion segmentation tasks.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"How is uncertainty handled for traffic speed prediction models?",{"text":80,"@type":76},"Two Gaussian process frameworks are proposed to quantify uncertainty, including a shallow GP equivalence of a CNN and a deep kernel CNN-GP with spatio-temporal kernels, evaluated using confidence intervals and noise analysis.",{"name":82,"@type":73,"acceptedAnswer":83},"How are cancer bone lesions localized in the proposed segmentation approach?",{"text":84,"@type":76},"An enhanced dataset is generated and a generative adversarial network reconstructs healthy and lesion-affected bones; lesion location is obtained by subtracting the original image from the reconstructed image.","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"]