原创力文档人工智能论文20220917 2019PhD Cambridge The resurgence of structure in deep neural networks_PVelickovic.pdfVIP
- 1、本文档被系统程序自动判定探测到侵权嫌疑，本站暂时做下架处理。
- 2、如果您确认为侵权，可联系本站左侧在线QQ客服请求删除。我们会保证在24小时内做出处理，应急电话：400-050-0827。
- 3、此文档由网友上传，因疑似侵权的原因，本站不提供该文档下载，只提供部分内容试读。如果您是出版社/作者，看到后可认领文档，您也可以联系本站进行批量认领。
Theresurgenceofstructure
indeepneuralnetworks
PetarVeliˇckovi´c
TrinityCollege
Thisdissertationissubmittedon15January2019
forthedegreeofDoctorofPhilosophy
Theresurgenceofstructureindeepneuralnetworks
PetarVeliˇckovi´c
Abstract
Machinelearningwithdeepneuralnetworks(“deeplearning”)allowsforlearningcom‐
plexfeaturesdirectlyfromrawinputdata,completelyeliminatinghand‐crafted,“hard‐
coded”featureextractionfromthelearningpipeline.Thishasleadtostate‐of‐the‐art
performancebeingachievedacrossseveral—previouslydisconnected—problemdomains,
includingcomputervision,naturallanguageprocessing,reinforcementlearningandgen‐
erativemodelling.Thesesuccessstoriesnearlyuniversallygohand‐in‐handwithavail‐
abilityofimmensequantitiesoflabelledtrainingexamples(“bigdata”)exhibitingsimple
grid‐likestructure(e.g.textorimages),exploitablethroughconvolutionalorrecurrent
layers.Thisisduetotheextremelylargenumberofdegrees‐of‐freedominneuralnet‐
works,leavingtheirgeneralisationabilityvulnerabletoeffectssuchasoverfitting.
However,thereremainmanydomainswhereextensivedatagatheringisnotalwaysap‐
propriate,affordable,orevenfeasible.Furthermore,dataisgenerallyorganisedinmore
complicatedkindsofstructure—whichmostexistingapproacheswouldsimplydiscard.
Examplesofsuchtasksareabundantinthebiomedicalspace;withe.g.smallnumbersof
subjectsavailableforanygivenclinicalstudy,orrelationshipsbetweenproteinsspecified
viainteractionnetworks.Ihypothesisethat,ifdeeplearningistoreachitsfullpotential
insuchenvironments,weneedtoreconsider“hard‐coded”approaches—integratingas‐
sumptionsaboutinherentstructureintheinputdatadirectlyintoourarchitecturesand
learningalgorithms,throughstructuralinductivebiases.Inthisdissertation,Idirectlyval‐
idate
您可能关注的文档
- 人工智能论文20230930 2023PhD MIT Towards General-purpose Vision via Multiview Contrastive Learning.pdf
- 人工智能论文20221120 2021PhD_Oxford_Identifying and Exploiting Structures for Reliable Deep Learning.pdf
- 人工智能论文20230211 2022PhD_Princeton_Bridging Theory and Practice in Deep Learning Optimization and Generalization.pdf
- 人工智能论文20240106 2023PhD Cambridge Learning from Structured Data with Weak Supervision.pdf
- 人工智能论文20231118 2023PhD MIT Multimodality Models, Algorithms, and Applications.pdf
- 人工智能论文20231112 2023PhD MIT Optimization Methods for Machine Learning under Structural Constraints.pdf
- 人工智能论文20230415 2023PhD Princeton Probabilistic models for structured biomedical data.pdf
- 人工智能论文20231021 2020PhD_Rochester_High-Capacity Directional Graph Networks.pdf
- 人工智能论文20231216 2023PhD ETH Zurich Neural Optimal Transport for Dynamical Systems Methods and Applications in Biomedicine.pdf
- 人工智能论文20231007 2023PhD UC Berkeley Geometry-Inspired Sampling Algorithms and Random Graphs.pdf
- 人工智能论文20220911 2021PhD HKUST Harnessing Graph Changes in Complex Graph Queries_Jian Xun.pdf
- 人工智能论文20220910 2022PhD Michigan Combining Supervised Dimension Reduction Techniques for High-Dimensional Data.pdf
- 人工智能论文20220828 2022PhD Cornell Combining At the Cutting Edge of Computational Pathology From Domain Knowledge Representation to Explorative Modelling.pdf
- 人工智能论文20220827 2022PhD Pisa Bayesian Deep Learning for Graphs_.pdf
- 人工智能论文20220806 2022PhD Northwestern Accelerating Materials Study and Discovery through Machine Learning.pdf
- 人工智能论文20220730 2022PhD Washington Combining Graph Techniques for Next Generation Cybersecurity.pdf
- 人工智能论文20220723 2022PhD Colorado Automating Investigative Pattern Detection Using Machine Learning & Graph Pattern Matching Techniques.pdf
- 人工智能论文20220716 2022PhD CMU Deep Learning Methods for Catalyst Surface and Interface Structure Analysis.pdf
- 人工智能论文20220709 2022PhD Arizona Mining Heterogeneous Spatial-Temporal Data with Graph Neural Network to Support Smart City Management.pdf
- 人工智能论文20220702 2022MS Windsor SEGCECO Subgraph Embedding of Gene Expression Matrix for Prediction of CEll-Cell COmmunication.pdf
文档评论（0）