Дисертації з теми "Exploration and exploitation of documents"
Оформте джерело за APA, MLA, Chicago, Harvard та іншими стилями
Ознайомтеся з топ-50 дисертацій для дослідження на тему "Exploration and exploitation of documents".
Біля кожної праці в переліку літератури доступна кнопка «Додати до бібліографії». Скористайтеся нею – і ми автоматично оформимо бібліографічне посилання на обрану працю в потрібному вам стилі цитування: APA, MLA, «Гарвард», «Чикаго», «Ванкувер» тощо.
Також ви можете завантажити повний текст наукової публікації у форматі «.pdf» та прочитати онлайн анотацію до роботи, якщо відповідні параметри наявні в метаданих.
Переглядайте дисертації для різних дисциплін та оформлюйте правильно вашу бібліографію.
Abeille, Marc. "Exploration-exploitation with Thompson sampling in linear systems." Thesis, Lille 1, 2017. http://www.theses.fr/2017LIL10182/document.
Повний текст джерелаThis dissertation is dedicated to the study of the Thompson Sampling (TS) algorithms designed to address the exploration-exploitation dilemma that is inherent in sequential decision-making under uncertainty. As opposed to algorithms derived from the optimism-in-the-face-of-uncertainty (OFU) principle, where the exploration is performed by selecting the most favorable model within the set of plausible one, TS algorithms rely on randomization to enhance the exploration, and thus are much more computationally efficient. We focus on linearly parametrized problems that allow for continuous state-action spaces, namely the Linear Bandit (LB) problems and the Linear Quadratic (LQ) control problems. We derive two novel analyses for the regret of TS algorithms in those settings. While the obtained regret bound for LB is similar to previous results, the proof sheds new light on the functioning of TS, and allows us to extend the analysis to LQ problems. As a result, we prove the first regret bound for TS in LQ, and show that the frequentist regret is of order O(sqrt{T}) which matches the existing guarantee for the regret of OFU algorithms in LQ. Finally, we propose an application of exploration-exploitation techniques to the practical problem of portfolio construction, and discuss the need for active exploration in this setting
Bazzazian, Navid. "Essays in Employee Entrepreneurship." Thesis, Jouy-en Josas, HEC, 2014. http://www.theses.fr/2014EHEC0015/document.
Повний текст джерелаThis dissertation consists of three essays on employee entrepreneurship where some employees leave paid employment to start their own business. In particular I examine the antecedents of such entrepreneurial transitions from contextual and individual perspectives. In the first essay I examine why historically best performing firms generate more entrepreneurs than other firms. In thesecond essay I investigate the relationship between matching in the labor market and entrepreneurial transition of employees. The empirical setting of the first two essays is the whole population of Swedish workers from 1990 to 2007. In the third essay I investigate the effect of organizational exploration and exploitation on employee entrepreneurship in medical devices industry. The overall conclusion of this dissertation is that context matters for entrepreneurship. In addition, sorting in the labor market and fit between an employer and employee determine who becomes an entrepreneur
Warlop, Romain. "Novel learning and exploration-exploitation methods for effective recommender systems." Thesis, Lille 1, 2018. http://www.theses.fr/2018LIL1I056/document.
Повний текст джерелаThis thesis, written in a company as a CIFRE thesis in the company fifty-five, studies recommender systems algorithms. We propose three new algorithms that improved over state-of-the-art solutions in terms of performance or matching industrial constraints. To that end, we proposed a first algorithm based on tensor factorization, a generalization of matrix factorization, commonly used on collaborative filtering. We then proposed a new algorithm that improves basket completion state-of-the-art algorithms. The goal of basket completion algorithms is to recommend a new product to a given user based on the products she is about to purchase in order to increase the user value. To that end we leverage Determinantal Point Processes, i.e., probability measure where the probability to observe a given set is proportional to the determinant of a kernel matrix. We generalized DPP approaches for basket completion using a tensor point of view coupled with a logistic regression. Finally, we proposed a reinforcement learning algorithm that allows to alternate between several recommender systems algorithms. Indeed, using always the same algorithm may either bore the user for a while or reinforce her trust in the system. Thus, the algorithm performance is not stationary and depends on when and how much the algorithm has been used in the past. Our reinforcement learning algorithm learns in real time how to alternate between several recommender system algorithms in order to maximize long term performances, that is in order to keep the user interested in the system as long as possible
Le, Glatin Mario. "Innover pour décider : Modéliser et expérimenter l'ambidextrie décisionnelle pour gérer les métabolismes de l'organisation innovante." Thesis, Paris Sciences et Lettres (ComUE), 2018. http://www.theses.fr/2018PSLEM039/document.
Повний текст джерелаIt is now common knowledge to develop an ambidextrous organization in a firm to guarantee a competitive advantage in its environment. Such perspective is rooted in James March’s model (1991) whose purpose is to sustain organizational learning through adaptive mechanisms between agents. This model anchored in the tradition of problem-solving offers a behavioural approach coping with biases and heuristics of bounded rationality. In contrast, with literature models studying the balance between exploration and exploitation activities (ambidexterity), we demonstrate how exploration can be more generative if it uses exploitation constraints as prior knowledge to generate concepts. These rely on positively conditioning them by decision-making parameters contributing to exploitation. New decision situations are designed instead of traditionally selecting explorations. This model - decisional ambidexterity-allows building up projects capable of exploring conditions overcoming limitations and performances of exploitation. A collaborative research conducted at Zodiac Aerospace allowed revealing such new model. It extends organizational ambidexterity into the unknown, more adapted to innovation management & governance, whilst precising how to drive organization metabolisms inducing change among participating organizations in a conglomerate of SMEs
Cordeil, Maxime. "Exploration et exploitation de l’espace de conception des transitions animées en visualisation d’information." Thesis, Toulouse, ISAE, 2013. http://www.theses.fr/2013ESAE0044/document.
Повний текст джерелаData visualizations allow information to be transmitted to users. In order to explore and understand the data, it is often necessary for users to manipulate the display of this data. When manipulating the visualization, visual transitions are necessary to avoid abrupt changes in this visualization, and to allow the user to focus on the graphical object of interest. These visual transitions can be coded as an animation, or techniques that link the data across several displays. The first aim of this thesis was to examine the benefits and properties of animated transitions used to explore and understand large quantities of multidimensional data. In order to do so, we created a taxonomy of existing animated transitions. This taxonomy allowed us to identify that no animated transition currently exists that allows the user to control the direction of objects during the transition. We therefore proposed an animated transition that allows the user to have this control during the animation. In addition, we studied an animated transition technique that uses 3D rotation to transition between visualizations. We identified the advantages of this technique and propose an improvement to the current design. The second objective was to study the visual transitions used in the Air Traffic Control domain. Air Traffic Controllers use a number of visualizations to view vast information which is duplicated in several places: the Radar screen, the strip board, airplane lists (departures/arrivals) etc. Air traffic controllers perform visual transitions as they search between these different displays of information. We studied the way animations can be used in the Air Traffic Control domain by implementing a radar image prototype which combines three visualizations typically used by Air Traffic Controllers
Jeon, Eui Ju. "Antecedents and Consequences of Exploration and Exploitation Decisions : Evidence from Corporate Venture Capital Investing." Thesis, Université Paris-Saclay (ComUE), 2017. http://www.theses.fr/2017SACLH001/document.
Повний текст джерелаThis dissertation addresses unexplored issues on the antecedents, management, and outcomes of corporate venture capital (CVC). More specifically, I examine how negative performance feedback and corporate governance influence the direction of organizational change ˗ in terms of exploration and exploitation ˗ and how balancing such change over time influences firm performance in the CVC context. I first review the extant literature on CVC and lay out the unique contributions of my research. Then, in the first essay, I theorize on how poor firm performance influences the resource allocation decisions on exploration and exploitation and how such decisions are affected by the concentration of dedicated and transient shareholders and by the board of directors' monitoring and advising intensities. In the second essay, I empirically examine how the resource allocation decisions on exploration and exploitation are influenced by dedicated and transient shareholders in the context of CVC investing. In the third essay, I examine how balancing exploration and exploitation over time and the characteristics of oscillation impact firm performance. The empirical analysis in the latter two essays is based on CVC investments made by 286 U.S. companies during 1993-2013. This dissertation contributes to the Behavioral Theory of the Firm and Corporate Governance research by introducing how shareholders and boards influence managerial decision-making in search and change, Ambidexterity research by studying how continuous change and organizational inertia impact temporal spillover between exploration and exploitation, and CVC research by examining the antecedents and consequences of explorative and exploitative initiatives in CVC investing
Salmeron, Jérémy. "Le processus d'innovation socio-économique : un processus ambidextre ?" Thesis, Lyon, 2016. http://www.theses.fr/2016LYSE3055/document.
Повний текст джерелаThis doctoral dissertation focus on the nature of innovation generated by the actors and on the dialectic management of exploitation and exploration phases among the socio-economic (SEAM) innovation process. The research is based on an intervention-research carried out during five years in a SME from computing industry. It aims at putting this innovation process is the field of ambidextrous innovation process. In the framework of a steered and instrumented organizational change process, this research is presenting the recycling of dysfunctions into innovation actions
Lepers, Thomas. "Identification et structuration des champs d'innovation dans une business unit." Thesis, Paris, CNAM, 2016. http://www.theses.fr/2016CNAM1068/document.
Повний текст джерелаOur case study based on an exploratory entity evolving from a small team to a business unit raise questions related to the identification and structuration of an innovation field. The literature gives few elements with regards to the choice of such a field, and the mechanisms to set priorities within a chosen field. Our results related to innovation processes show that exploratory activities leads to a renewal of exploitation, and also that the difficulties encountered by NCA in exploration activities are less related to the existing processes than to the choice of an innovation field. Our results related to organization show that NCA works in a mode of simultaneous ambidexterity, being able at the same time to launch new product categories and to explore new product categories. This also leads to the renewal of existing routines. The structure of NCA has also some original features, being at the same time a business unit and an exploratory entity. This exploratory business unit seems to be able to launch more easily new products categories, because there is no need to look for a landing zone, as it is the case of organisations solely responsible for exploration
Toussaint, Maxime. "Exploitation et exploration de la diversité génétique d’une population naturelle de Streptomyces issue d’un micro-habitat sol." Thesis, Université de Lorraine, 2018. http://www.theses.fr/2018LORR0027/document.
Повний текст джерелаStreptomyces are known to possess a large enzymatic arsenal which can have important roles in the soil. During this thesis, we explored their genetic, functional and ecological diversity using collections from forest soils. Thus, the exploration of their cellulolytic potential and their ability to detect complex sugars released by wood during lignivorous fungi attacks has led to the creation of a biosensor whose exploitation could constitute a new normative tool for the detection of the degradation of wood. Subsequent to comparative genomic approach carried out between sympatric isolates, our results also demonstrated that phylogenetically highly related strains exhibited large differences in the presence / absence of genes, suggesting a rapid rate of evolution of the population accessory genome. These genes, often associated with potentially transferable elements, underlined important role of horizontal transfer for population diversification. Using a reverse ecology approach, the predicted function of some of these genes could also be correlated with a potential ecological role. Thus, one of the variable gene clusters identified by genome analysis was involved in the production of secondary metabolites and would constitute a common good for the population. All of our results confirm the wide metabolic diversity of Streptomyces (and their utility for applied purposes), but also indicates that this diversification would be rapid between nearby strains and would have an important ecological role in the natural populations of Streptomyces
Kacioui-Maurin, Elodie. "Exploration et exploitation, les formes d'ambidextrie organisationnelle dans l'innovation de services : le cas des prestataires de services logistiques." Thesis, Aix-Marseille 2, 2011. http://www.theses.fr/2011AIX24013/document.
Повний текст джерелаThe research that we carry out examines the organization of incremental and radical innovations, both intra- and inter-enterprise, in the light of organizational ambidextry works (Tushman and O’Reilly, 1996, 1999; O’Reilly and Tushman, 2004; Birkinshaw and Gibson, 2004).Taking the case of logistics service provision, an industry concerned by the extensive innovation dilemma (Le Masson, Weil and Hatchuel, 2006), our research aims to study the organizational modalities implemented to innovate, but also to identify and understand the explanatory factors of the recourse to organizational ambidextry forms. Moreover, to our knowledge, no study has been performed on organizational ambidextry in the service industry. It is therefore, also about the identification of the specificities bound to the organization of innovation on service companies.To successfully achieve our research, case studies were performed with 8 Logistics Service Providers (LSP) companies. 49 interviews were carried out with the actors participating in the innovation process in each company. Secondary data (internal documentation, professional press, etc.) was also collected.The results enlighten the organizational modalities implemented to innovate in service companies and within interconnected companies for both researchers and practitioners. Our study also identifies the explanatory factors of the recourse to, and the combination of, organizational ambidextry forms. These factors are bound to the company’s characteristics as well as the characteristics of the innovation, but also to the inter-organizational relationship between the provider and its costumer
Letessier, Pierre. "Découverte et exploitation d'objets visuels fréquents dans des collections multimédia." Thesis, Paris, ENST, 2013. http://www.theses.fr/2013ENST0014/document.
Повний текст джерелаThe main goal of this thesis is to discover frequent visual objects in large multimedia collections. As in many areas (finance, genetics, . . .), it consists in extracting a knowledge, using the occurence frequency of an object in a collection as a relevance criterion. A first contribution is to provide a formalism to the problems of mining and discovery of frequent visual objects. The second contribution is a generic method to solve these two problems, based on an iterative sampling process, and on an efficient and scalable rigid objects matching. The third contribution of this work focuses on building a likelihood function close to the perfect distribution. Experiments show that contrary to state-of-the-art methods, our approach allows to discover efficiently very small objects in several millions images. Finally, several applications are presented, including trademark logos discovery, transmedia events detection or visual-based query suggestion
Cogliati, Dezza Irene. "“Vanilla, Vanilla .but what about Pistachio?” A Computational Cognitive Clinical Neuroscience Approach to the Exploration-Exploitation Dilemma." Doctoral thesis, Universite Libre de Bruxelles, 2018. https://dipot.ulb.ac.be/dspace/bitstream/2013/278730/3/Document1.pdf.
Повний текст джерелаDoctorat en Sciences psychologiques et de l'éducation
info:eu-repo/semantics/nonPublished
Mbarki, Mohamed. "Gestion de l'hétérogénéité documentaire : le cas d'un entrepôt de documents multimédia." Toulouse 3, 2008. http://thesesups.ups-tlse.fr/185/.
Повний текст джерелаThe knowledge society is based on three axes: the diffusion and use of information via new technologies, the deduction of knowledge induced by this information and the economic impacts which can result from this information. To offer to the actors and more particularly to the "decision makers" of this society some tools which enable them to produce and manage "knowledge" or at least "elements of knowledge" seem to be rather difficult to ensure. This difficulty is due to the dynamism of the environment and the diversity of factors influencing the information production, extraction and communication. Indeed, this information is included in documents which are collected from disseminated sources (Internet, Workflow, numerical libraries, etc. ). These documents are thus heterogeneous on the content and on the form (they can be related to various fields, they can be more or less structured, they can have various structures, they contain several type of media, are stored in several type of supports, etc). The current challenges are to conceive new applications to exploit this document heterogeneity. Having in mind these needs, the work presented in my thesis, aims to face these challenges and in particular at proposing solutions in order "to manage and create knowledge" starting from the integration of all information available on the heterogeneous documents. The handling of multimedia documents repositories constitutes the applicative framework of our proposals. Our approach is articulated around three complementary axes: (1) the representation, (2) storage (or integration) and (3) exploitation of the heterogeneous documents. Documents representation is related to the determination of information that must be preserved and the way according to which they must be organized to offer better apprehending and envisaging of their uses. The solution that we chose to meet these needs bases on the proposal for a documents model which integrates several overlapping and complementary levels of description (a generic layer and a specific one, a logical description and a semantic one). .
Gutehrlé, Nicolas. "Information extraction from unstructured documents for the valorisation of historical periodicals : application to the heritage of the Bourgogne Franche-Comté Region in France." Electronic Thesis or Diss., Bourgogne Franche-Comté, 2024. http://www.theses.fr/2024UBFCC006.
Повний текст джерелаIn recent years, libraries and archives have undertaken numerous digitisation campaigns to widen public access to their archival collections. However, the challenge of promoting the content of collections and making these resources discoverable remains. Digitisation often produces unstructured content that is difficult to navigate, while interfaces that rely on keyword-based queries to access archival materials can provide users with irrelevant results. In order to exploit the potential of the "Big Data of the Past", notion introduced by Kaplan and di Lenardo in 2017, it is essential to develop methods and frameworks for structuring the textual content of documents, with the aim of improving their exploration and discoverability. In this context, the present Ph.D. thesis addresses the problem of processing digitised historical documents, focusing on the extraction of Named Entities and Relations in order to create interfaces for the efficient exploitation of historical textual data. Firstly, we propose a new method for determining the logical structure of historical newspapers using a rule-based approach. Secondly, we present a method for extracting entities and relations about people and places mentioned in texts. Our approach is called Extensible, Lightweight and Interpretable Joint Extraction of Relations and Entities (ELIJERE). It is based on linguistic resources obtained through distant supervision. Finally, we propose a general framework for studying the expression of spatial information in documents, and another framework for applying TimeLine Summarisation methods to document collections. We show how these methods can be applied to produce semantically rich interfaces, such as timelines and maps, that allow the general public a close and a distant reading of these collections
Laquitaine, Steeve. "Les bases neuronales de l’apprentissage décisionnel au sein des ganglions de la base : étude électrophysiologique et comportementale chez le primate non humain." Thesis, Bordeaux 2, 2010. http://www.theses.fr/2010BOR21743/document.
Повний текст джерелаA fundamental question in neuroscience, as well as in various fields such as economics, psychology and sociology, concerns the decision making processes by which animals and humans select actions based on reward and punishment. Both decision making processes and their neural basis are still poorly understood. Also, both human and animals often make suboptimal decisions in many tasks studied. Our first aim is to improve the understanding of why such sub-optimal decisions are made. Also, the alteration of decision making processes causes diseases, the understanding of whose mechanisms is essential in developing better treatment strategies. In this report, we propose a new approach which consists in extracting the neural substrates of choice behavior heterogeneity in between sessions. Our results show that although primates learn on average to identify the best option and converge to an optimal policy in a consequent number of sessions, they fail on average to optimize their behavior. We revealed that this suboptimal behavior was characterized by an unexpected high behavioral heterogeneity during the task that was due to the creation of irrelevant preferences by the monkeys. We finally show that although a few neurons of the putamen encode the action value, their contribution to the overall population activity is weak. Putamen activity rather reflects the futures performances and predicts the creation of rational and irrational behaviors
Nevers, Yannis Alain. "Exploitation de marqueurs évolutifs pour l'étude des relations génotype-phénotype : application aux ciliopathies." Thesis, Strasbourg, 2018. http://www.theses.fr/2018STRAJ090/document.
Повний текст джерелаIn the omics era, the study of genotype-phenotype relations requires the integration of a wide variety of data to describe diverse aspects of biological systems. Comparative genomics provides an original perspective, that of evolution, allowing the exploitation of the wide phenotypic diversity of living species. My thesis focused on the design of evolutionary markers to describe genes according to their evolutionary history. First, I built an exhaustive orthology resource, called OrthoInspector 3.0, to extract synthetic evolutionary information from genomic data. I then developed methods to explore the markers in relation to functional or phenotypic data. These methods have been incorporated in the OrthoInspector resource, as well as in the MyGeneFriends social network and applied to the study of ciliopathies, leading to the identification of 87 new ciliary genes
Aklil, Nassim. "Apprentissage actif sous contrainte de budget en robotique et en neurosciences computationnelles. Localisation robotique et modélisation comportementale en environnement non stationnaire." Thesis, Paris 6, 2017. http://www.theses.fr/2017PA066225/document.
Повний текст джерелаDecision-making is a highly researched field in science, be it in neuroscience to understand the processes underlying animal decision-making, or in robotics to model efficient and rapid decision-making processes in real environments. In neuroscience, this problem is resolved online with sequential decision-making models based on reinforcement learning. In robotics, the primary objective is efficiency, in order to be deployed in real environments. However, in robotics what can be called the budget and which concerns the limitations inherent to the hardware, such as computation times, limited actions available to the robot or the lifetime of the robot battery, are often not taken into account at the present time. We propose in this thesis to introduce the notion of budget as an explicit constraint in the robotic learning processes applied to a localization task by implementing a model based on work developed in statistical learning that processes data under explicit constraints, limiting the input of data or imposing a more explicit time constraint. In order to discuss an online functioning of this type of budgeted learning algorithms, we also discuss some possible inspirations that could be taken on the side of computational neuroscience. In this context, the alternation between information retrieval for location and the decision to move for a robot may be indirectly linked to the notion of exploration-exploitation compromise. We present our contribution to the modeling of this compromise in animals in a non-stationary task involving different levels of uncertainty, and we make the link with the methods of multi-armed bandits
Epure, Elena Viorica. "Modélisation automatique des conversations en tant que processus d'intentions de discours interdépendantes." Thesis, Paris 1, 2018. http://www.theses.fr/2018PA01E068/document.
Повний текст джерелаThe proliferation of digital data has enabled scientific and practitioner communities to createnew data-driven technologies to learn about user behaviors in order to deliver better services and support to people in their digital experience. The majority of these technologies extensively derive value from data logs passively generated during the human-computer interaction. A particularity of these behavioral traces is that they are structured. However, the pro-actively generated text across Internet is highly unstructured and represents the overwhelming majority of behavioral traces. To date, despite its prevalence and the relevance of behavioral knowledge to many domains, such as recommender systems, cyber-security and social network analysis,the digital text is still insufficiently tackled as traces of human behavior to automatically reveal extensive insights into behavior.The main objective of this thesis is to propose a corpus-independent method to automatically exploit the asynchronous communication as pro-actively generated behavior traces in order to discover process models of conversations, centered on comprehensive speech intentions and relations. The solution is built in three iterations, following a design science approach.Multiple original contributions are made. The only systematic study to date on the automatic modeling of asynchronous communication with speech intentions is conducted. A speech intention taxonomy is derived from linguistics to model the asynchronous communication and, comparedto all taxonomies from the related works, it is corpus-independent, comprehensive—as in both finer-grained and exhaustive in the given context, and its application by non-experts is proven feasible through extensive experiments. A corpus-independent, automatic method to annotate utterances of asynchronous communication with the proposed speech intention taxonomy is designed based on supervised machine learning. For this, validated ground-truth corpora arecreated and groups of features—discourse, content and conversation-related, are engineered to be used by the classifiers. In particular, some of the discourse features are novel and defined by considering linguistic means to express speech intentions, without relying on the corpus explicit content, domain or on specificities of the asynchronous communication types. Then, an automatic method based on process mining is designed to generate process models of interrelated speech intentions from conversation turns, annotated with multiple speech intentions per sentence. As process mining relies on well-defined structured event logs, an algorithm to produce such logs from conversations is proposed. Additionally, an extensive design rationale on how conversations annotated with multiple labels per sentence could be transformed in event logs and what is the impact of different decisions on the output behavioral models is released to support future research. Experiments and qualitative validations in medicine and conversation analysis show that the proposed solution reveals reliable and relevant results, but also limitations are identified,to be addressed in future works
Galichet, Nicolas. "Contributions to Multi-Armed Bandits : Risk-Awareness and Sub-Sampling for Linear Contextual Bandits." Thesis, Paris 11, 2015. http://www.theses.fr/2015PA112242/document.
Повний текст джерелаThis thesis focuses on sequential decision making in unknown environment, and more particularly on the Multi-Armed Bandit (MAB) setting, defined by Lai and Robbins in the 50s. During the last decade, many theoretical and algorithmic studies have been aimed at cthe exploration vs exploitation tradeoff at the core of MABs, where Exploitation is biased toward the best options visited so far while Exploration is biased toward options rarely visited, to enforce the discovery of the the true best choices. MAB applications range from medicine (the elicitation of the best prescriptions) to e-commerce (recommendations, advertisements) and optimal policies (e.g., in the energy domain). The contributions presented in this dissertation tackle the exploration vs exploitation dilemma under two angles. The first contribution is centered on risk avoidance. Exploration in unknown environments often has adverse effects: for instance exploratory trajectories of a robot can entail physical damages for the robot or its environment. We thus define the exploration vs exploitation vs safety (EES) tradeoff, and propose three new algorithms addressing the EES dilemma. Firstly and under strong assumptions, the MIN algorithm provides a robust behavior with guarantees of logarithmic regret, matching the state of the art with a high robustness w.r.t. hyper-parameter setting (as opposed to, e.g. UCB (Auer 2002)). Secondly, the MARAB algorithm aims at optimizing the cumulative 'Conditional Value at Risk' (CVar) rewards, originated from the economics domain, with excellent empirical performances compared to (Sani et al. 2012), though without any theoretical guarantees. Finally, the MARABOUT algorithm modifies the CVar estimation and yields both theoretical guarantees and a good empirical behavior. The second contribution concerns the contextual bandit setting, where additional informations are provided to support the decision making, such as the user details in the ontent recommendation domain, or the patient history in the medical domain. The study focuses on how to make a choice between two arms with different numbers of samples. Traditionally, a confidence region is derived for each arm based on the associated samples, and the 'Optimism in front of the unknown' principle implements the choice of the arm with maximal upper confidence bound. An alternative, pioneered by (Baransi et al. 2014), and called BESA, proceeds instead by subsampling without replacement the larger sample set. In this framework, we designed a contextual bandit algorithm based on sub-sampling without replacement, relaxing the (unrealistic) assumption that all arm reward distributions rely on the same parameter. The CL-BESA algorithm yields both theoretical guarantees of logarithmic regret and good empirical behavior
Atwi, Aliaa. "Exploration vs. exploitation in coupon personalization." Thesis, Massachusetts Institute of Technology, 2018. http://hdl.handle.net/1721.1/115729.
Повний текст джерелаCataloged from student-submitted PDF version of thesis.
Includes bibliographical references (pages 51-52).
Personalized offers aim to maximize profit by taking into account customer preferences inferred from past purchase behavior. For large retailers with extensive product offerings, learning customer preferences can be challenging due to relatively short purchase histories of most customers. To alleviate the dearth of data, we propose exploiting similarities among products and among customers to reduce problem dimensions. We also propose that retailers use personalized offers not only to maximize expected profit, but to actively learn their customers' preferences. An offer that does not maximize expected profit given current information may still provide valuable insights about customer preferences. This information enables more profitable coupon allocation and higher profits in the long run. In this thesis we 1) derive approximate inference algorithms to learn customer preferences from purchase data in real time, 2) formulate the retailers' offer allocation problem as a multi armed bandit and explore solution strategies.
by Aliaa Atwi.
Elec. E. in Computer Science
Bouhlel, Imen. "Essais sur le dilemme exploration-exploitation." Thesis, Université Côte d'Azur (ComUE), 2019. http://theses.univ-cotedazur.fr/2019AZUR0037.
Повний текст джерелаA growing body of empirical evidence during the two last decades has been showing inconsistencies between individual choices when the individuals make decisions from description (i.e., when they are provided with a perfect knowledge about the states space, including all the possible outcomes, and the underlying probabilities), compared to when they make decisions from experience (i.e., when they do not know all the possible outcomes or/and their occurrence probabilities). These inconsistencies are referred to as the description/experience gap. Undersearch has been pointed out as one of the key determinants of this gap. Hence, even though little studied in economics, search becomes a central question, deserving serious interest. This thesis aims at contributing to the theoretical and experimental literature studying search and the related exploration-exploitation dilemma, both at the individual and at the collective level. The thesis is made of 3 essays, combining theoretical, agent-based modelling, evolutionary simulations and laboratory experiments. The first chapter of this thesis examines the determinants of search behavior in the context of an individual optimal stopping problem and shows that this behavior largely depends on the degree of certainty of the information, and is affected by both regret and anticipation. The second chapter investigates information sharing behavior in competitive collective search using agent-based and evolutionary simulations. It finds robust evidence for the individual benefits of sharing, even when others do not reciprocate, as long as two mechanisms as present: Imitation with a certain level of innovation and local visibility. The third chapter experimentally tests and supports the validity of theses results, and stresses the crucial role of learning
Adelsbo, Felix. "Exploration and Exploitation in Reinforcement Learning." Thesis, KTH, Skolan för teknikvetenskap (SCI), 2018. http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-235841.
Повний текст джерелаInom förstärkande inlärning existerar ett dilemma inom utforskningkontra utnyttjande. Detta har lett till utvecklandet av metoder som har olika tillvägagångssätt för detta. Att använda olika typer av metoder, och modifiera dem på olika sätt kan leda till olika resultat. Kunskap om hur olika metoder fungerar kan ge kunskap om vad som ska användas i ett specifikt fall.Två sätt som metoder kan modifieras är ändring i justerbara parametrar och ändring i antalet slumpmässiga steg i början. Hur mycket dessa två modifieringar påverkar resultatet i en specifik miljö kan skilja sig mycketåt, och kan vara en väldigt kritisk sak att betänka för ett visst resultat.Målet med denna studie är att besvara frågan om hur prestandan på de olika metoderna random, greedy, e- greedy e-decreasing och Softmax påverkas av olika värden på deras justerbara parametrar, och av antaletslumpmässiga steg i början. Simuleringarna och en jämförande analys utförs för fallet med en inverterad pendel med en vertikal stolpe placerad på en rörlig vagn.
Ozcan, Ozkan. "Balancing exploration and exploitation in agent learning." Thesis, Monterey, California. Naval Postgraduate School, 2011. http://hdl.handle.net/10945/5468.
Повний текст джерелаControlling the ratio of exploration and exploitation in agent learning in dynamic environments is a continuing challenge in applying agent-learning techniques. Methods to control this ratio in a manner that mimics human behavior are required for use in the representation of human behavior in simulations, where the goal is to constrain agent-learning mechanisms in a manner similar to that observed in human cognition. The Cultural Geography (CG) model, under development in TRAC Monterey, is an agent-based social simulation. It simulates a wide variety of situations and scenarios so that a dynamic ratio between exploration and exploitation makes the decisions more sensible. As part of an attempt to improve the model, this thesis investigates enhancements to the exploration-exploitation balance by using different techniques. The work includes design of experiments with a range of factors in multiple environments and statistical analysis related to these experiments. As a main finding from this research, for small environments and for short runs techniques based on subjective utility give better results, while for long runs techniques based on time obtain higher utilities than other techniques. In more complex and bigger environments, a combined technique performed better in long runs.
Lampela, T. (Teemu). "Modelling exploration and exploitation in organizational learning." Bachelor's thesis, University of Oulu, 2019. http://jultika.oulu.fi/Record/nbnfioulu-201904271553.
Повний текст джерелаAhsan, Nasir. "Combining Exploration and Exploitation in Active Learning." Thesis, The University of Sydney, 2013. http://hdl.handle.net/2123/10144.
Повний текст джерелаDichio, Vito. "The exploration-exploitation paradigm : a biophysical approach." Electronic Thesis or Diss., Sorbonne université, 2023. http://www.theses.fr/2023SORUS402.
Повний текст джерелаThe study of living systems is notoriously challenging. The often-quoted daunting complexity of biological systems is primarily due to the intricacies of their interactions, their multiple organisation levels and their dynamic nature. In the quest to understand this complexity, parallels drawn with standard physics – in particular, statistical physics -- are both useful and of limited use. On the one hand, they provide a rich set of theoretical and methodological building blocks for constructing theories and designing experiments. On the other hand, life also unfolds according to principles that are unparalleled in the physics of conventional matter. A crucial difference lies in the notion of function: biological systems are shaped by the need to perform specific tasks. A general problem for living systems is to find and promote those configurations that yield improved or optimal functions, we call this the exploration-exploitation (EE) problem. One specific instance of the above is found in evolutionary biology. There, random genetic mutations sustain the exploration of the configuration space, with those leading to higher reproductive success being favoured by natural selection. Inspired by the latter, we develop a novel formalism that encodes a general exploration-exploitation dynamics for biological networks. In particular, our EE dynamics is represented as an exploration of a functional landscape and consists of stochastic configuration changes combined with the state-dependent optimisation of an objective function (F metric). We begin by investigating its main features through the study of simple, analytically tractable functional landscapes. We deploy simulations for more general and complex applications. We then turn to the brain wiring problem, i.e., the development of an individual's nervous system during its early life. We argue that this is another specific instance of the EE problem and therefore can be addressed by using our theoretical framework. In particular, we focus on brain maturation in the nematode C.elegans, the only organism for which a complete network of neurons and neuronal connections has been reconstructed, at multiple developmental time points (seven). We fix the network at birth and use the adult stage to infer (i) a parsimonious maxent (ERG) description of the F metric for the worm brain and (ii) the two parameters of our EE dynamics. According to the topography of its functional landscape, the adult brain is characterised by a tendency to form both triads and high degree nodes. We demonstrate that our EE dynamics in such landscape is capable of tracking down the entire developmental history. In particular, we show that the trajectory we obtain closely reproduces the other experimental time points that we did not use for inference. This is true both in the space of model statistics and for a number of other network properties. Additionally, we discuss a micro-level interpretation of the EE dynamics in terms of the underlying synapse formation process. Our study is a first step towards the system-level understanding of the development of a natural brain and can be extended (i) to encompass more complex functional landscapes, (ii) to different organisms than the C. elegans and (iii) to several different problems than the brain wiring. Indeed, we posit that the exploration-exploitation paradigm is among those life-specific principles that we are just beginning to uncover
Liedholm, Johnson Eva. "Mineral Rights : Legal Systems Governing Exploration and Exploitation." Doctoral thesis, KTH, Fastighetsvetenskap, 2010. http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-12044.
Повний текст джерелаQC20100723
Shaposhnik, Yaron. "Exploration vs. exploitation : reducing uncertainty in operational problems." Thesis, Massachusetts Institute of Technology, 2016. http://hdl.handle.net/1721.1/106681.
Повний текст джерелаThis electronic version was submitted by the student author. The certified thesis is available in the Institute Archives and Special Collections.
Cataloged from student-submitted PDF version of thesis.
Includes bibliographical references (pages 205-207).
Motivated by several core operational applications, we introduce a class of multistage stochastic optimization models that capture a fundamental tradeoff between performing work under uncertainty (exploitation) and investing resources to reduce the uncertainty in the decision making (exploration/testing). Unlike existing models, in which the exploration-exploitation tradeoffs typically relate to learning the underlying distributions, the models we introduce assume a known probabilistic characterization of the uncertainty, and focus on the tradeoff of learning exact realizations. In the first part of the thesis (Chapter 2), we study a class of scheduling problems that capture common settings in service environments in which the service provider must serve a collection of jobs that have a-priori uncertain processing times and priorities (modeled as weights). In addition, the service provider must decide how to dynamically allocate capacity between processing jobs and testing jobs to learn more about their respective processing times and weights. We obtain structural results of optimal policies that provide managerial insights, efficient optimal and near-optimal algorithms, and quantification of the value of testing. In the second part of the thesis (Chapter 3), we generalize the model introduced in the first part by studying how to prioritize testing when jobs have different uncertainties. We model difference in uncertainties using the convex order, a general relation between distributions, which implies that the variance of one distribution is higher than the variance of the other distribution. Using an analysis based on the concept of mean preserving local spread, we show that the structure of the optimal policy generalizes that of the initial model where jobs were homogeneous and had equal weights. Finally, in the third part of the thesis (Chapter 4), we study a broad class of stochastic combinatorial optimization that can be formulated as Linear Programs whose objective coefficients are random variables that can be tested, and whose constraint polyhedron has the structure of a polymatroid. We characterize the optimal policy and show that similar types of policies optimally govern testing decisions in this setting as well.
by Yaron Shaposhnik.
Ph. D.
Edwards, James. "Exploration and exploitation in Bayes sequential decision problems." Thesis, Lancaster University, 2016. http://eprints.lancs.ac.uk/84589/.
Повний текст джерелаCottini, Ciro <1975>. "The complexity of exploration and exploitation in organizations." Doctoral thesis, Alma Mater Studiorum - Università di Bologna, 2020. http://amsdottorato.unibo.it/9325/11/Cottini_The_complexity_of_exploration_and_exploitation_in_organizations_090320.pdf.
Повний текст джерелаPuhan, Tatjana-Xenia. "Balancing exploration and exploitation by creating organizational think tanks." Wiesbaden Betriebswirtschaftlicher Verl. Gabler, 2007. http://d-nb.info/987457306/04.
Повний текст джерелаGolder, Peter Thomas. "Exploitation and exploration of advanced technology in financial institutions." Thesis, City University London, 1997. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.389547.
Повний текст джерелаBailey, Jennifer. "Synchronizing exploration and exploitation: knowledge creation challenges in innovation." Diss., Georgia Institute of Technology, 2013. http://hdl.handle.net/1853/50378.
Повний текст джерелаHess, Michael. "Network Frontier: Reframing Exploration and Exploitation in Internet Rhetoric." Thesis, University of Oregon, 2015. http://hdl.handle.net/1794/19198.
Повний текст джерелаBell, Alexander. "Studies on exploration and exploitation : concepts, roles and dynamics." Thesis, Paris Sciences et Lettres (ComUE), 2017. http://www.theses.fr/2017PSLED077.
Повний текст джерелаCompetition and technological change have never been as intense as they are today. To survive, companies must increasingly carry out current activities at the same time that they explore future ones. It is no easy task to operate on both of these fronts, and doing so is a potential source of tension and contradiction. Although management research has provided insight into the exploration and exploitation tension, there are still contradictions and unexplored areas. In particular, there is no consensus on the definition of the concepts of exploration and exploitation, which means the research lacks a certain level of consistency. In addition, there has been very little attention paid to the Exploration/Exploitation tension in the context of young, growing “scale-up” companies, which are particularly exposed to threats to their survival. To obtain a better understanding of the situations this tension causes in these companies, we have adopted a multi-paper dissertation consisting of three complementary studies.In the first study, to create a conceptual framework for the thesis, we answer researchers’ call for a clarification of the concepts of exploration and exploitation by proposing a typology of exploration. Then, based on a longitudinal, multi-case approach focused on eight scale-up companies, we first explore, in a second study, the factors influencing firms’ strategic orientations vis-à-vis exploration and exploitation activities. Finally, in the third study, we seek to understand how management teams define their individual roles with respect to exploration and exploitation activities
Pickering, Andrew Christopher. "An empirical analysis of the exploitation of oil." Thesis, University of Exeter, 2002. http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.248158.
Повний текст джерелаDebarbieux, Denis. "Modélisation et requêtes des documents semi-structurés : exploitation de la structure de graphe." Phd thesis, Université des Sciences et Technologie de Lille - Lille I, 2005. http://tel.archives-ouvertes.fr/tel-00619303.
Повний текст джерелаGoffin, Pascal. "An Exploration of Word-Scale Visualizations for Text Documents." Thesis, Université Paris-Saclay (ComUE), 2016. http://www.theses.fr/2016SACLS256/document.
Повний текст джерелаThis dissertation explores how embedding small data-driven contextual visualizations can complement text documents. More specifically, I identify and define important aspects and relevant research directions for the integration of small data-driven contextual visualizations into text. This integration should eventually become as fluid as writing and as usable as reading a text. I define word-scale visualisations as small data-driven contextual visualizations embedded in text documents. These visualizations can use various visual encodings including geographical maps, heat maps, pie charts, and more complex visualizations. They can appear at a range of word scales, including sizes larger than a letter, but smaller than a sentence or paragraph. Word-scale visualisations can help support and be used in many forms of written discourse such as text books, notes, blog posts, reports, stories, or poems. As graphical supplements to text, word-scale visualisations can be used to emphasize certain elements of a document (e.g. a word or a sentence), or to provide additional information. For example, a small stock chart can be embedded next to the name of a company to provide additional information about the past trends of its stocks. In another example, game statistics can be embedded next to the names of soccer teams or players in daily reports from the UEFA European Championship. These word-scale visualisations can then for example allow readers to make comparison between number of passes of teams and players. The main benefit of word-scale visualisations is that the reader can remain focused on the text as the visualization are within the text rather than alongside it.In the thesis, I make the following main contributions: I explore why word-scale visualisations can be useful and how to support their creation. I investigate placement options to embed word-scale visualisations and quantify their effects on the layout and flow of the text. As word-scale visualisations also have implications on the reader's reading behavior I propose a first study that investigates different word-scale visualisation positions on the reading behavior. I also explore how word-scale visualisations can be combined with interaction to support a more active reading by proposing interaction methods to collect, arrange and compare word-scale visualisations. Finally, I propose design considerations for the authoring of word-scale visualisations and conclude with application examples.In summary, this dissertation contributes to the understanding of small data-driven contextual visualizations embedded into text and their value for Information Visualization
Schläpfer, Jörg. "Exploitation vs. Exploration Organization Development in the European Insurance Industry /." St. Gallen, 2006. http://www.biblio.unisg.ch/org/biblio/edoc.nsf/wwwDisplayIdentifier/03601648001/$FILE/03601648001.pdf.
Повний текст джерелаAsree, Susita. "Challenges in the Global Supply Chain: Exploitation versus Exploration Strategy." Toledo, Ohio : University of Toledo, 2010. http://rave.ohiolink.edu/etdc/view?acc%5Fnum=toledo1271874448.
Повний текст джерелаTypescript. "Submitted to the Graduate Faculty as partial fulfillment of the requirements for The Doctor of Philosophy in Manufacturing Management and Engineering." "A dissertation entitled"--at head of title. Title from title page of PDF document. Bibliography: p. 158-180.
Johnston, Iain G. "Exploration, exploitation & complexity in biological evolution and self-assembly." Thesis, University of Oxford, 2010. http://ora.ox.ac.uk/objects/uuid:9dbf2fdc-219b-48dd-8bb7-2c6031070dc1.
Повний текст джерелаLeks, Klaudia, and Valquiria Jablinski. "The transition from exploitation to exploration in young entrepreneurial firms." Thesis, Uppsala universitet, Företagsekonomiska institutionen, 2018. http://urn.kb.se/resolve?urn=urn:nbn:se:uu:diva-354062.
Повний текст джерелаXiao, Song. "Balancing exploration and exploitation in robust multiobjective electromagnetic design optimisation." Thesis, University of Southampton, 2014. https://eprints.soton.ac.uk/370598/.
Повний текст джерелаLeloup, Benoît. "L'incertitude de deuxième ordre en économie : le compromis "exploration-exploitation"." Cachan, Ecole normale supérieure, 2002. http://www.theses.fr/2002DENS0013.
Повний текст джерелаBooth, Ramie. "Exploring the paradox of managerial ambidexterity in exploitation versus exploration." Diss., University of Pretoria, 2017. http://hdl.handle.net/2263/64875.
Повний текст джерелаMini Dissertation (MBA)--University of Pretoria, 2017.
pa2018
Gordon Institute of Business Science (GIBS)
MBA
Unrestricted
Hayashi, Shogo. "Information Exploration and Exploitation for Machine Learning with Small Data." Doctoral thesis, Kyoto University, 2021. http://hdl.handle.net/2433/263774.
Повний текст джерелаSykulski, Adam M. "The exploration-exploitation trade-off in sequential decision making problems." Thesis, Imperial College London, 2011. http://hdl.handle.net/10044/1/9073.
Повний текст джерелаEshtiwi, Suaad. "Analyse de la langue et exploitation pédagogique des documents authentiques en classe de langue." Thesis, Artois, 2018. http://www.theses.fr/2018ARTO0004.
Повний текст джерелаIn Lybia, French as a foreign language intervenes as a optional language, and only at the university level. Therefore, the training undertaken in Lybian universities - especially at Zawyia University- doesn’t allow students to acquire the main skills to face any kind of communication situations, especially the ones that have to do with daily life, because the university’s French department suffers from the lack of material and resources, also because teachers frequently didn’t received continuous training in French as a foreign language’s didactic, and finally because the methods concerned with the transmission of information don’t arouse leaners’ motivation. While thinking how to improve its teaching, how to teach differently a foreign language, how to make a foreign language alive in the classroom and lead to a more efficient learning, we decided to choose the use of authentic documents, which – to our opinion – can contribute to the modernization of French teaching methods in Lybia and develop language capabilities of learners. Consequently, we have been piloting the dependability of the introduction of this methodology in the French as a foreign language department of the University of Zawyia and testing its effectiveness. This experiment took place throughout the year 2017 with learners in second year of study of French as a foreign language with A2+/B1 level. In our mind, and based on the learners’ results, the integration turned out to be convincing because it arouses interest and motivation, allows to vary teaching aids of this foreign language, to break with the standard practices of teaching and because it develops language, communicative, and intercultural skills for learners
Faroque, Anisur Rahman. "Network exploration and exploitation in international entrepreneurship: an opportunity-based view." Thesis, University of Canterbury. Department of Management, Marketing & Entrepreneurship, 2014. http://hdl.handle.net/10092/9639.
Повний текст джерелаSmith, Gregory P. "Morphological charts a systematic exploration of qualitative design space /." Connect to this title online, 2007. http://etd.lib.clemson.edu/documents/1202500458/.
Повний текст джерела