Rozprawy doktorskie na temat „Active learning”
Utwórz poprawne odniesienie w stylach APA, MLA, Chicago, Harvard i wielu innych
Sprawdź 50 najlepszych rozpraw doktorskich naukowych na temat „Active learning”.
Przycisk „Dodaj do bibliografii” jest dostępny obok każdej pracy w bibliografii. Użyj go – a my automatycznie utworzymy odniesienie bibliograficzne do wybranej pracy w stylu cytowania, którego potrzebujesz: APA, MLA, Harvard, Chicago, Vancouver itp.
Możesz również pobrać pełny tekst publikacji naukowej w formacie „.pdf” i przeczytać adnotację do pracy online, jeśli odpowiednie parametry są dostępne w metadanych.
Przeglądaj rozprawy doktorskie z różnych dziedzin i twórz odpowiednie bibliografie.
Hsu, Daniel Joseph. "Algorithms for active learning". Diss., [La Jolla] : University of California, San Diego, 2010. http://wwwlib.umi.com/cr/ucsd/fullcit?p3404377.
Pełny tekst źródłaTitle from first page of PDF file (viewed June 10, 2010). Available via ProQuest Digital Dissertations. Vita. Includes bibliographical references (leaves 97-101).
Simpson, Colin Gordon. "Exploring Chinese business management students' experience of active learning pedagogies : how much action is possible in active learning classrooms?" Thesis, University of Exeter, 2013. http://hdl.handle.net/10871/14660.
Pełny tekst źródłaBrinker, Klaus. "Active learning with kernel machines". [S.l. : s.n.], 2004. http://deposit.ddb.de/cgi-bin/dokserv?idn=974403946.
Pełny tekst źródłaRibeiro, de Mello Carlos Eduardo. "Active Learning : an unbiased approach". Phd thesis, Châtenay-Malabry, Ecole centrale de Paris, 2013. http://tel.archives-ouvertes.fr/tel-01000266.
Pełny tekst źródłaZhao, Liyue. "Active Learning with Unreliable Annotations". Doctoral diss., University of Central Florida, 2013. http://digital.library.ucf.edu/cdm/ref/collection/ETD/id/5893.
Pełny tekst źródłaPh.D.
Doctorate
Computer Science
Engineering and Computer Science
Computer Science
Ganti, Mahapatruni Ravi Sastry. "New formulations for active learning". Diss., Georgia Institute of Technology, 2014. http://hdl.handle.net/1853/51801.
Pełny tekst źródłaHuddy, Vyvyan. "Active processing in implicit learning". Thesis, University of Glasgow, 2001. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.390696.
Pełny tekst źródłaWilliams, Kevin. "Active Learning for drug discovery". Thesis, Aberystwyth University, 2014. http://hdl.handle.net/2160/eaf6e66e-17fe-41a9-ac1d-9939abbb8331.
Pełny tekst źródłaMohamad, Saad. "Active learning for data streams". Thesis, Bournemouth University, 2017. http://eprints.bournemouth.ac.uk/29901/.
Pełny tekst źródłaLima, Vinicius Gomes de. "Peer effects in active learning". reponame:Repositório Institucional do FGV, 2017. http://hdl.handle.net/10438/18273.
Pełny tekst źródłaApproved for entry into archive by Suzinei Teles Garcia Garcia (suzinei.garcia@fgv.br) on 2017-05-19T18:36:54Z (GMT) No. of bitstreams: 1 dissertacao_viniciuslima.pdf: 333215 bytes, checksum: db37308bb8e35440208daa9ef10ee9ba (MD5)
Made available in DSpace on 2017-05-19T18:40:11Z (GMT). No. of bitstreams: 1 dissertacao_viniciuslima.pdf: 333215 bytes, checksum: db37308bb8e35440208daa9ef10ee9ba (MD5) Previous issue date: 2017-04-24
This paper investigates peer effects in higher education in an environment of active learning that gives great importance for students’ interaction through group work. Our empirical strategy uses exogenous variation in group composition to estimate peer effects in different exercises. We find no evidence of peer effects in a basic linear-in-means specification considering all assigned peers. However, we find positive and statistically significant impact of peers coming from student’s same high school. We also find no evidence of peer effects with a model that takes into account student and peers’ position in the ability distribution.
Este trabalho investiga efeito de pares no ensino superior em um ambiente de active learning que dá grande importância à interação dos estudantes através do trabalho em grupo. A estratégia empírica utiliza variação exógena na composição dos grupos para estimar o efeito dos pares em diferentes exercícios. Não encontramos evidência de efeito de pares numa especificação linear-in-means básica considerando todos os pares do grupo atribuído ao aluno. Entretanto, encontramos efeito positivo e estatisticamente significante de pares que frequentaram a mesma escola de ensino médio. Não encontramos evidência de efeito de pares em um modelos que procura considerar a posição dos alunos na distribuição de habilidade.
Monteleoni, Claire E. (Claire Elizabeth) 1975. "Learning with online constraints : shifting concepts and active learning". Thesis, Massachusetts Institute of Technology, 2006. http://hdl.handle.net/1721.1/38308.
Pełny tekst źródłaIncludes bibliographical references (p. 99-102).
Many practical problems such as forecasting, real-time decision making, streaming data applications, and resource-constrained learning, can be modeled as learning with online constraints. This thesis is concerned with analyzing and designing algorithms for learning under the following online constraints: i) The algorithm has only sequential, or one-at-time, access to data. ii) The time and space complexity of the algorithm must not scale with the number of observations. We analyze learning with online constraints in a variety of settings, including active learning. The active learning model is applicable to any domain in which unlabeled data is easy to come by and there exists a (potentially difficult or expensive) mechanism by which to attain labels. First, we analyze a supervised learning framework in which no statistical assumptions are made about the sequence of observations, and algorithms are evaluated based on their regret, i.e. their relative prediction loss with respect to the hindsight-optimal algorithm in a comparator class. We derive a, lower bound on regret for a class of online learning algorithms designed to track shifting concepts in this framework. We apply an algorithm we provided in previous work, that avoids this lower bound, to an energy-management problem in wireless networks, and demonstrate this application in a network simulation.
(cont.) Second, we analyze a supervised learning framework in which the observations are assumed to be iid, and algorithms are compared by the number of prediction mistakes made in reaching a target generalization error. We provide a lower bound on mistakes for Perceptron, a standard online learning algorithm, for this framework. We introduce a modification to Perceptron and show that it avoids this lower bound, and in fact attains the optimal mistake-complexity for this setting. Third, we motivate and analyze an online active learning framework. The observations are assumed to be iid, and algorithms are judged by the number of label queries to reach a target generalization error. Our lower bound applies to the active learning setting as well, as a lower bound on labels for Perceptron paired with any active learning rule. We provide a new online active learning algorithm that avoids the lower bound, and we upper bound its label-complexity. The upper bound is optimal and also bounds the algorithm's total errors (labeled and unlabeled). We analyze the algorithm further, yielding a label-complexity bound under relaxed assumptions. Using optical character recognition data, we empirically compare the new algorithm to an online active learning algorithm with data-dependent performance guarantees, as well as to the combined variants of these two algorithms.
by Claire E. Monteleoni.
Ph.D.
Hu, François. "Semi-supervised learning in insurance : fairness and active learning". Electronic Thesis or Diss., Institut polytechnique de Paris, 2022. http://www.theses.fr/2022IPPAG002.
Pełny tekst źródłaInsurance organisations store voluminous textual data sources on a daily basis (free text fields used by telephonists, emails, customer reviews, ...). However, this mass of textual data involves specific issues in terms of regulations, such as compliance with the privacy constraints imposed in Europe by the recent General Data Protection Regulation (GDPR) : this textual data may contain information that is not compliant with the RGPD standards, thus raising ethical issues and cannot be retained by the insurer. Today, this textual data is tagged by experts (oracles) and this process is not suitable for managing large volumes and near real-time information. Therefore, the implementation of an accurate (in terms of prediction), low-cost (in terms of labelling) and ethical (in terms of fairness) learning system is needed in insurance and this thesis addresses and solves some of these challenges. The first challenge is to reduce the labelling effort (thus focusing on data quality) with the help of active learning, a feedback loop between model inference and an oracle: since in insurance unlabelled data is usually abundant, active learning can become an important asset to reduce the cost of labelling. Another major challenge is the issue of fairness in Machine Learning model inferences. Since inequalities and discriminations can be found in the data, learning models are likely to reproduce some unfairness, making them unusable in production. This thesis explores these problems and proposes solutions, especially for multi-class classification tasks. In particular, we propose an algorithmic fairness method that guarantees either exact fairness at the expense of model accuracy, or a compromise between fairness and accuracy called epsilon-fairness. In addition, we propose a fair active learning method that requests informative instances while making the model fair. The proposed methodologies have the advantage of being agnostic with respect to the statistical learning model. These results are studied and applied on real and synthetic datasets
Barnabé-Lortie, Vincent. "Active Learning for One-class Classification". Thesis, Université d'Ottawa / University of Ottawa, 2015. http://hdl.handle.net/10393/33001.
Pełny tekst źródłaMuhivuwomunda, Divine. "Data De-Duplication through Active Learning". Thesis, University of Ottawa (Canada), 2010. http://hdl.handle.net/10393/28859.
Pełny tekst źródłaSchurr, Jochen. "On assortment optimization under active learning". Thesis, Lancaster University, 2012. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.658059.
Pełny tekst źródłaJohannemann, Jonathan. "COAL : a continuous active learning system". Thesis, Massachusetts Institute of Technology, 2017. http://hdl.handle.net/1721.1/111453.
Pełny tekst źródłaCataloged from PDF version of thesis.
Includes bibliographical references (pages 59-60).
In this thesis, our objective is to enable businesses looking to enhance their product by varying its attributes, where effectiveness of the new product is assessed by humans. To achieve this, we mapped the task to a machine learning problem. The solution is two fold: learn a non linear model that can map the attribute space to the human response, which can then be used to make predictions, and an active learning strategy that enables learning this model incrementally. We developed a system called Continuous active learning system (COAL).
by Jonathan Johannemann.
M. Fin.
Budnik, Mateusz. "Active and deep learning for multimedia". Thesis, Université Grenoble Alpes (ComUE), 2017. http://www.theses.fr/2017GREAM011.
Pełny tekst źródłaThe main topics of this thesis include the use of active learning-based methods and deep learning in the context of retrieval of multimodal documents. The contributions proposed during this thesis address both these topics. An active learning framework was introduced, which allows for a more efficient annotation of broadcast TV videos thanks to the propagation of labels, the use of multimodal data and selection strategies. Several different scenarios and experiments were considered in the context of person identification in videos, including using different modalities (such as faces, speech segments and overlaid text) and different selection strategies. The whole system was additionally validated in a dry run involving real human annotators.A second major contribution was the investigation and use of deep learning (in particular the convolutional neural network) for video retrieval. A comprehensive study was made using different neural network architectures and training techniques such as fine-tuning or using separate classifiers like SVM. A comparison was made between learned features (the output of neural networks) and engineered features. Despite the lower performance of the engineered features, fusion between these two types of features increases overall performance.Finally, the use of convolutional neural network for speaker identification using spectrograms is explored. The results are compared to other state-of-the-art speaker identification systems. Different fusion approaches are also tested. The proposed approach obtains comparable results to some of the other tested approaches and offers an increase in performance when fused with the output of the best system
Chernets, M. "Teaching strategies to promote active learning". Thesis, Київський національний університет технологій та дизайну, 2019. https://er.knutd.edu.ua/handle/123456789/13027.
Pełny tekst źródłaДорда, Світлана Володимирівна, Светлана Владимировна Дорда i Svitlana Volodymyrivna Dorda. "Using the WEB for active learning". Thesis, Sumy State University, 2003. http://essuir.sumdu.edu.ua/handle/123456789/62784.
Pełny tekst źródłaHříbek, David. "Active Learning pro zpracování archivních pramenů". Master's thesis, Vysoké učení technické v Brně. Fakulta informačních technologií, 2021. http://www.nusl.cz/ntk/nusl-445535.
Pełny tekst źródłaWeir, Jennifer Anne. "Active learning in transportation engineering education". Link to electronic thesis, 2004. http://www.wpi.edu/Pubs/ETD/Available/etd-12214-155616/.
Pełny tekst źródłaTaylor, Teresa Brooks. "Being Intentional: Active Learning, Student Reflection". Digital Commons @ East Tennessee State University, 2000. https://dc.etsu.edu/etsu-works/3645.
Pełny tekst źródłaKihlström, Helena. "Active Stereo Reconstruction using Deep Learning". Thesis, Linköpings universitet, Institutionen för medicinsk teknik, 2019. http://urn.kb.se/resolve?urn=urn:nbn:se:liu:diva-158276.
Pełny tekst źródłaSusag, Angie. "Expeditionary learning". Diss., [Missoula, Mont.] : The University of Montana, 2009. http://etd.lib.umt.edu/theses/available/etd-07142009-111349.
Pełny tekst źródłaGür, Hülya. "Learning to teach mathematics and the place of active learning". Thesis, University of Leicester, 1999. http://hdl.handle.net/2381/30939.
Pełny tekst źródłaAlabdulrahman, Rabaa. "A Comparative Study of Ensemble Active Learning". Thesis, Université d'Ottawa / University of Ottawa, 2014. http://hdl.handle.net/10393/31805.
Pełny tekst źródłaEvans, Cindy. "The Effect of Faculty Development on Active Learning in the College Classroom". Thesis, University of North Texas, 2001. https://digital.library.unt.edu/ark:/67531/metadc2762/.
Pełny tekst źródłaScott, Fiona Marie. "Action-reflection-learning in a lean production environment /". St. Lucia, Qld, 2002. http://www.library.uq.edu.au/pdfserve.php?image=thesisabs/absthe17167.pdf.
Pełny tekst źródłaBerlind, Christopher. "New insights on the power of active learning". Diss., Georgia Institute of Technology, 2015. http://hdl.handle.net/1853/53948.
Pełny tekst źródłaDenton, Stephen E. "Exploring active learning in a Bayesian framework". [Bloomington, Ind.] : Indiana University, 2009. http://gateway.proquest.com/openurl?url_ver=Z39.88-2004&rft_val_fmt=info:ofi/fmt:kev:mtx:dissertation&res_dat=xri:pqdiss&rft_dat=xri:pqdiss:3380073.
Pełny tekst źródłaTitle from PDF t.p. (viewed on Jul 19, 2010). Source: Dissertation Abstracts International, Volume: 70-12, Section: B, page: 7870. Advisers: John K. Kruschke; Jerome R. Busemeyer.
Houlsby, Neil. "Efficient Bayesian active learning and matrix modelling". Thesis, University of Cambridge, 2014. https://www.repository.cam.ac.uk/handle/1810/248885.
Pełny tekst źródłaZhao, Wenquan. "Deep Active Learning for Short-Text Classification". Thesis, KTH, Skolan för datavetenskap och kommunikation (CSC), 2017. http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-212577.
Pełny tekst źródłaI detta arbete studerar vi en ny aktiv inlärningsalgoritm som appliceras på en djup inlärningsarkitektur för klassificering av korta (kinesiska) texter. Ämnesområdet hör därmedtill ett ämnesöverskridande område mellan aktiv inlärning och inlärning i djupa nätverk .En av flaskhalsarna i djupa nätverk när de används för klassificering är att de beror avtillgången på många klassificerade datapunkter. Dessa är dyra och tidskrävande att skapa. Aktiv inlärning syftar till att överkomma denna typ av nackdel genom att generera frågor rörande de mest informativa oklassade datapunkterna och få dessa klassificerade. Aktiv inlärning syftar med andra ord till att uppnå bästa klassificeringsprestanda medanvändandet av så få klassificerade datapunkter som möjligt. Denna idé har studeratsinom konventionell maskininlärning, som tex supportvektormaskinen (SVM) för bildklassificering samt inom djupa neuronnätverk inkluderande bl.a. convolutional networks(CNN) och djupa beliefnetworks (DBN) för bildklassificering. Emellertid är kombinationenav aktiv inlärning och rekurrenta nätverk (RNNs) för klassificering av korta textersällsynt. Vi demonstrerar här resultat för klassificering av korta texter ur en databas frånZhuiyi Inc. Att notera är att för att uppnå bättre klassificeringsnoggranhet med lägre beräkningsarbete (overhead) så uppvisar den föreslagna algoritmen stora minskningar i detantal klassificerade träningspunkter som behövs jämfört med användandet av slumpvisadatapunkter. Vidare, den föreslagna algoritmen är något bättre än den konventionellaurvalsmetoden, osäkherhetsurval (uncertanty sampling). Den föreslagna aktiva inlärningsalgoritmen minska dramatiskt den mängd klassificerade datapunkter utan att signifikant påverka klassificeringsnoggranheten hos den ursprungliga RNN-klassificeraren när den tränats på hela datamängden. För några fall uppnår den föreslagna algoritmen t.o.m.bättre klassificeringsnoggranhet än denna ursprungliga RNN-klassificerare.
Cora, Vlad M. "Model-based active learning in hierarchical policies". Thesis, University of British Columbia, 2008. http://hdl.handle.net/2429/737.
Pełny tekst źródłaWatanabe, Yukio. "Learning control of automotive active suspension systems". Thesis, Cranfield University, 1997. http://dspace.lib.cranfield.ac.uk/handle/1826/13865.
Pełny tekst źródłaBheda, Anuj. "Predictive analytics of active learning based education". Thesis, Massachusetts Institute of Technology, 2017. http://hdl.handle.net/1721.1/113509.
Pełny tekst źródłaCataloged from PDF version of thesis.
Includes bibliographical references (pages 113-115).
Learning Analytics (LA) is defined as the collection, measurement, and analysis of data related to student performance such that the feedback from the analytical insights can be used to optimize student learning and improve student outcomes. Blended Learning (BL) is a teaching paradigm that involves a mix of face-to-face interactions in a classroom based setting along with instructional material distributed through an online medium. In this thesis, we explore the role of a blended learning model coupled with learning analytics in an introductory programming class for non-computer science students. We identify the features that were necessary for setting up the infrastructure of the course. These include discussions on preparing the course content materials and producing assignment exercises. We then talk about the various dynamics that were in play during the duration of the class by describing the interplay between watching video tutorials, listening to mini-lectures and performing active learning exercises that are backed by modern software development practices. Lastly, we spend time analyzing the data collected to create a predictive model that can measure student performance by defining the specifications of a machine learning algorithm along with many of its adjustable parameters. The system thus created will allow instructors to identify possible outliers in teaching efficacy, the feedback from which could then be used to tune course material for the betterment of student outcomes.
by Anuj Bheda.
S.M. in Engineering and Management
Shepherdson, Emma 1972. "Teaching concepts utilizing active learning computer environments". Thesis, Massachusetts Institute of Technology, 2001. http://hdl.handle.net/1721.1/84215.
Pełny tekst źródłaHansson, Kim, i Erik Hörlin. "Active learning via Transduction in Regression Forests". Thesis, Blekinge Tekniska Högskola, Institutionen för kreativa teknologier, 2015. http://urn.kb.se/resolve?urn=urn:nbn:se:bth-10935.
Pełny tekst źródłaVan, Amerom W. P. C. "Active / interactive learning facilitation in large classes". Journal for New Generation Sciences, Vol 3, Issue 2: Central University of Technology, Free State, Bloemfontein, 2005. http://hdl.handle.net/11462/483.
Pełny tekst źródłaAnyone who has taught a large class is aware of the physical and emotional constraints upon both lecturer and students. For students the dominant problems are anonymity, passivity and a frustration of not being able to say what is happening to them. For lecturers the dominant problems are not being able to relate to students as individuals, a feeling of being driven back to traditional teaching, being overwhelmed by assessment demands, and a sense of not being in control of the class. An increase in class size requires lecturers radically to reconsider how they deliver their courses. One such strategy proposed in this paper is that of active learning facilitation - getting students to work and think in the classroom about what and why they are doing what they are doing.
Tan, Run Yan. "Active Learning using a Sample Selector Network". Thesis, KTH, Skolan för elektroteknik och datavetenskap (EECS), 2020. http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-287312.
Pełny tekst źródłaI detta arbete sätter vi steget i en begränsad märkningsbudget och föreslår att vi använder ett provväljarnätverk för att lära och välja effektiva träningsprover, vars etiketter vi sedan skulle skaffa för att träna målmodellen som utför den nödvändiga maskininlärningsuppgiften. Vi antar att provfunktionerna, tillståndet för målmodellen och utbildningsförlusten för målmodellen är informativa för att träna provväljarnätverket. Dessutom uppskattar vi målmodellens tillstånd med dess mellanliggande och slutliga nätverksutgångar. Vi undersöker om provväljarnätverket enligt en begränsad märkningsbudget kan lära sig och välja utbildningsprover som tränar målmodellen minst lika effektivt som att använda en annan träningsdel av samma storlek som är enhetligt slumpmässigt samplad från hela utbildningsdatasystemet, det senare är det vanliga förfarandet som används för att utbilda maskininlärningsmodeller utan aktivt lärande. Vi hänvisar till denna vanliga procedur som den traditionella maskininlärning enhetliga slumpmässig sampling metod. Vi utför experiment på datasätten MNIST och CIFAR-10; och visa med empiriska bevis att under en begränsad märkningsbudget och vissa andra förhållanden, aktivt lärande med hjälp av ett provvalnätverk gör det möjligt för målmodellen att lära sig mer effektivt.
Ahsan, Nasir. "Combining Exploration and Exploitation in Active Learning". Thesis, The University of Sydney, 2013. http://hdl.handle.net/2123/10144.
Pełny tekst źródłaArens, Robert Segre Alberto Maria. "Learning to rank documents with support vector machines via active learning". [Iowa City, Iowa] : University of Iowa, 2009. http://ir.uiowa.edu/etd/331.
Pełny tekst źródłaNainabasti, Binod. "Role of Students’ Participation on Learning Physics in Active Learning Classes". FIU Digital Commons, 2016. http://digitalcommons.fiu.edu/etd/3022.
Pełny tekst źródłaArens, Robert James. "Learning to rank documents with support vector machines via active learning". Diss., University of Iowa, 2009. https://ir.uiowa.edu/etd/331.
Pełny tekst źródłaChen, Si. "Active Learning Under Limited Interaction with Data Labeler". Thesis, Virginia Tech, 2021. http://hdl.handle.net/10919/104894.
Pełny tekst źródłaM.S.
Machine Learning (ML) has achieved huge success in recent years. Machine Learning technologies such as recommendation system, speech recognition and image recognition play an important role on human daily life. This success mainly build upon the use of large amount of labeled data: Compared with traditional programming, a ML algorithm does not rely on explicit instructions from human; instead, it takes the data along with the label as input, and aims to learn a function that can correctly map data to the label space by itself. However, data labeling requires human effort and could be time-consuming and expensive especially for datasets that contain domain-specific knowledge (e.g., disease prediction etc.) Active Learning (AL) is one of the solution to reduce data labeling effort. Specifically, the learning algorithm actively selects data points that provide more information for the model, hence a better model can be achieved with less labeled data. While traditional AL strategies do achieve good performance, it requires a small amount of labeled data as initialization and performs data selection in multi-round, which pose great challenge to its application, as there is no platform provide timely online interaction with data labeler and the interaction is often time inefficient. To deal with the limitations, we first propose DULO which a new setting of AL is studied: data selection is only allowed to be performed once. To further broaden the application of our method, we propose D²ULO which is built upon DULO and Domain Adaptation techniques to avoid the use of initial labeled data. Our experiments show that both of the proposed two frameworks achieve better performance compared with state-of-the-art baselines.
Swiatocha, Andrea Leigh. "Learning through Movement". Thesis, Virginia Tech, 2013. http://hdl.handle.net/10919/51847.
Pełny tekst źródłaMaster of Architecture
Bloodgood, Michael. "Active learning with support vector machines for imbalanced datasets and a method for stopping active learning based on stabilizing predictions". Access to citation, abstract and download form provided by ProQuest Information and Learning Company; downloadable PDF file, 200 p, 2009. http://proquest.umi.com/pqdweb?did=1818417671&sid=1&Fmt=2&clientId=8331&RQT=309&VName=PQD.
Pełny tekst źródłaCalma, Adrian [Verfasser]. "Active learning with uncertain annotators : towards dedicated collaborative interactive learning / Adrian Calma". Kassel : kassel university press c/o Universität Kassel - Universitätsbibliothek, 2020. http://d-nb.info/1230907955/34.
Pełny tekst źródłaHuang, Jian Giles C. Lee. "A multiclass boosting classification method with active learning". [University Park, Pa.] : Pennsylvania State University, 2009. http://etda.libraries.psu.edu/theses/approved/WorldWideIndex/ETD-4765/index.html.
Pełny tekst źródłaHasenjäger, Martina. "Active data selection in supervised and unsupervised learning". [S.l. : s.n.], 2000. http://deposit.ddb.de/cgi-bin/dokserv?idn=960209220.
Pełny tekst źródłaJaulmes, Robin. "Active learning in partially observable Markov decision processes". Thesis, McGill University, 2006. http://digitool.Library.McGill.CA:80/R/?func=dbin-jump-full&object_id=98733.
Pełny tekst źródłaOur goal is to build Artificial Intelligence algorithms able to reproduce the reasoning of humans for these complex problems. We use the Reinforcement Learning framework, which allows to learn optimal behaviors in dynamic environments. More precisely, we adapt Partially-Observable Markov Decision Processes (POMDPs) to environments that are partially known.
We take inspiration from the field of Active Learning: we assume the existence of an oracle, who can, during a short learning phase, provide the agent with additional information about its environment. The agent actively learns everything that is useful in the environment, with a minimum use of the oracle.
After reviewing existing methods for solving learning problems in partially observable environments, we expose a theoretical active learning setup. We propose an algorithm, MEDUSA, and show theoretical and empirical proofs of performance for it.