Global ETD Search

11	Strategic and environmental uncertainty in social dilemmas Lindahl, Therese January 2005 (has links) Social dilemmas constitute a broad class of quandaries, including, for example, common pool resource (CPR) dilemmas and public good (PG) dilemmas. CPR's are characterized by non-excludability and rivalry and are often associated with overexploitation. Through similar arguments, the features non-excludability and non-rivalry give rise to under-provision of PG's. The prevalence and inefficiencies often associated with CPR's have given rise to an extensive literature and the role of resource uncertainty has not been ignored. Uncertainty combined with rivalry is often said to augment users' incentive to overexploit. However, underlying most of the theoretical research is an explicit or implicit assumption of symmetric information, or a symmetric lack of information. In reality, people generally have access to different sources of information and they may differ in their abilities to process information. In the first two papers of this thesis, the assumption of symmetry is relaxed and both papers demonstrate that from a welfare perspective, the distribution of uncertainty is also of importance. Many CPR's and PG's are natural, which can complicate the situation. In the traditional resource management literature, the exploited resource is often assumed to be properly characterized by some concave growth function. Today, there is extensive empirical evidence suggesting that many ecosystems have more complex, often non-linear dynamics. Management of such resources can be quite challenging as the non-linear dynamics can make the ecosystem flip between alternate stable states, and even marginal changes can cause radical transformations of such ecosystems. Most of the CPR models assume the shared resource to be of fixed size or to be able to generate a constant flow of services. In the third paper we aim at providing a more complete picture of the overexploitation of a common resource, by combining the institutional structure with complex ecological dynamics. We manage to raise questions and doubts about the standard assumptions. Another feature of convex-concave resources is that a state can become highly robust and sometimes an ecosystem change may even be irreversible. This is problematic if, for example, we wish to restore a degraded ecosystem. The aim of the fourth paper is to empirically analyze this question, by eliciting peoples' preferences through a hypothetical referendum on the issue. / Diss. Stockholm : Handelshögskolan, 2005 Common pool resources Asymmetric information Experiment Fairness Ecosystem complexity Convex-concave resources Grasslands Differential game Overgrazing Baltic Sea Hypothetical referendum Scope test Irreversibility Socialekologi-- matematiska modeller Naturresurser Miljöpolitik Miljöekonomi Economics Nationalekonomi
12	Essays on interest rate theory Elhouar, Mikael January 2008 (has links) Diss. (sammanfattning) Stockholm : Handelshögskolan, 2008 Sammanfattning jämte 3 uppsatser Interest rate models Regime switching Markovian realizations Expectations hypothesis Heath-Jarrow-Morton Swap market model State space models expectations hypotheses LIBOR market model Ränta matematiska modeller Business and economics Ekonomi
13	Revision of an artificial neural network enabling industrial sorting Malmgren, Henrik January 2019 (has links) Convolutional artificial neural networks can be applied for image-based object classification to inform automated actions, such as handling of objects on a production line. The present thesis describes theoretical background for creating a classifier and explores the effects of introducing a set of relatively recent techniques to an existing ensemble of classifiers in use for an industrial sorting system.The findings indicate that it's important to use spatial variety dropout regularization for high resolution image inputs, and use an optimizer configuration with good convergence properties. The findings also demonstrate examples of ensemble classifiers being effectively consolidated into unified models using the distillation technique. An analogue arrangement with optimization against multiple output targets, incorporating additional information, showed accuracy gains comparable to ensembling. For use of the classifier on test data with statistics different than those of the dataset, results indicate that augmentation of the input data during classifier creation helps performance, but would, in the current case, likely need to be guided by information about the distribution shift to have sufficiently positive impact to enable a practical application. I suggest, for future development, updated architectures, automated hyperparameter search and leveraging the bountiful unlabeled data potentially available from production lines. artificial neural networks machine learning deep learning connectionism pattern recognition machine learning automation image analysis information technology applied mathematics mathematical optimization information theory mathematical statistics mathematical models stochastic models probabilities chance approximations algorithms computer programs computer software signal processing high performance computing numerical methods high technology industries sustainable development artificiella neurala nätverk maskininlärning djup maskininlärning konnektionism mönsterigenkänning automatisering bildanalys informationsteknik tillämpad matematik optimering informationsteori statistisk inferens matematiska modeller stokastiska modeller sannolikhetskalkyl slumpen approximationer algoritmer datorprogram programvara signalbehandling högpresterande beräkningar numeriska metoder teknikutveckling maskinindustri högteknologisk industri maskinhandel skrothandel bärkraftig utveckling Robotics Robotteknik och automation Probability Theory and Statistics Sannolikhetsteori och statistik Other Computer and Information Science Annan data- och informationsvetenskap Signal Processing Signalbehandling

Page generated in 0.0637 seconds