• Refine Query
  • Source
  • Publication year
  • to
  • Language
  • 15
  • 1
  • 1
  • 1
  • Tagged with
  • 26
  • 8
  • 7
  • 6
  • 5
  • 5
  • 5
  • 4
  • 3
  • 3
  • 3
  • 3
  • 3
  • 3
  • 3
  • About
  • The Global ETD Search service is a free service for researchers to find electronic theses and dissertations. This service is provided by the Networked Digital Library of Theses and Dissertations.
    Our metadata is collected from universities around the world. If you manage a university/consortium/country archive and want to be added, details can be found on the NDLTD website.
21

Souběžný evoluční návrh hardwaru a softwaru / Concurrent evolutionary design of hardware and software

Minařík, Miloš January 2018 (has links)
Genetické programování (GP) je v určitém rozsahu schopno automaticky generovat požadované programy, aniž by uživatel musel určit, jakým způsobem má program postupovat. GP bylo s úspěchem použito k řešení široké škály praktických problémů z různých oblastí, přičemž výsledky byly často srovnatelné s řešeními vytvořenými člověkem. Doposud však nebyla zodpovězena otázka, zda GP dokáže generovat vysoce optimalizovaný výpočetní model (platformu) spolu s programem spustitelným na této platformě, který by řešil daný problém při dodržení všech omezení (například na plochu na čipu a zpoždění). V případě scénářů, kdy je optimalizováno více kritérií, by uživatelským výstupem měla být množina nedominovaných řešení s různými kombinacemi úrovně využití zdrojů (plocha, příkon) a výkonu (rychlosti provádění). Tento problém může být chápán jako souběžný návrh hardwaru a softwaru, zkráceně HW/SW codesign. Tato práce zkoumá způsoby, jakými lze souběžně evolučně vyvíjet platformu a programy v případě, že je problém zadán množinou vektorů vstupů a jim odpovídajících výstupů. Nejprve byl vytvořen model architektury a evoluční platforma zajišťující zpracování a evoluční vývoj těchto architektur. Kandidátní mikroprogramové architektury byly evolvovány spolu s programy pomocí lineárního genetického programování. Následně byla provedena série jednodušších experimentů. Navržená platforma dosahovala výsledků srovnatelných s nejnovějšími metodami. Na základě slabých míst objevených během počátečních experimentů byla platforma rozšířena. Rozšířená platforma byla poté ověřena na několika složitějších experimentech. Jeden z nich byla zaměřen na efektivní implementaci aproximace sigmoidální funkce. Platforma v tomto případě našla řadu různých řešení implementujících aproximaci sigmoidy, z nichž některá byla sekvenční a jiná čistě kombinační. V rámci experimentu byly evolučně nalezeny i známé algoritmy, přičemž některé z nich byly evolucí dokonce optimalizovány pro podmnožinu definičního oboru zvolenou pro daný experiment. Poslední sada experimentů byla zaměřena na evoluční návrh obrazových filtrů pro redukci šumu typu sůl a pepř. Platforma v tomto případě znovuobjevila koncept přepínaných filtrů a naezla variantu přepínaného mediánového filtru, která byla z hlediska výsledků filtrace srovnatelná s běžně používanými metodami. Tato práce prokázala, že pomocí genetického programování lze navrhovat a optimalizovat malé HW/SW systémy. Automatizovaný evoluční návrh složitějších HW/SW systémů zůstává otevřeným problémem vhodným k dalšímu výzkumu.
22

GENERATIVE MODELS WITH MARGINAL CONSTRAINTS

Bingjing Tang (16380291) 16 June 2023 (has links)
<p> Generative models form powerful tools for learning data distributions and simulating new samples. Recent years have seen significant advances in the flexibility and applicability of such models, with Bayesian approaches like nonparametric Bayesian models and deep neural network models such as Variational Autoencoders (VAEs) and Generative Adversarial Networks (GANs) finding use in a wide range of domains. However, the black-box nature of these models means that they are often hard to interpret, and they often come with modeling implications that are inconsistent with side knowledge resulting from domain knowledge. This thesis studies situations where the modeler has side knowledge represented as probability distributions on functionals of the objects being modeled, and we study methods to incorporate this particular kind of side knowledge into flexible generative models. This dissertation covers three main parts. </p> <p><br></p> <p>The first part focuses on incorporating a special case of the aforementioned side knowledge into flexible nonparametric Bayesian models. Many times, practitioners have additional distributional information about a subset of the coordinates of the observations being modeled. The flexibility of nonparametric Bayesian models usually implies incompatibility with this side information. Such inconsistency triggers the necessity of developing methods to incorporate this side knowledge into flexible nonparametric Bayesian models. We design a specialized generative process to build in this side knowledge and propose a novel sigmoid Gaussian process conditional model. We also develop a corresponding posterior sampling method based on data augmentation to overcome a doubly intractable problem. We illustrate the efficacy of our proposed constrained nonparametric Bayesian model in a variety of real-world scenarios including modeling environmental and earthquake data. </p> <p><br></p> <p>The second part of the dissertation discusses neural network approaches to satisfying the said general side knowledge. Further, the generative models considered in this part broaden into black-box models. We formulate this side knowledge incorporation problem as a constrained divergence minimization problem and propose two scalable neural network approaches as its solution. We demonstrate their practicality using various synthetic and real examples. </p> <p><br></p> <p> The third part of the dissertation concentrates on a specific generative model of individual pixels of the fMRI data constructed from a latent group image. Usually there is two-fold side knowledge about the latent group image: spatial structure and partial activation zones. The former can be captured by modeling the prior for the group image with Markov random fields. The latter, which is often obtained from previous related studies, is left for future research. We propose a novel Bayesian model with Markov random fields and aim to estimate the maximum a posteriori for the group image. We also derive a variational Bayes algorithm to overcome local optima in the optimization.</p>
23

Asymptomatic Recurrent Spontaneous Pneumoperitoneum

Faruqi, S A., Joshi, P N., Haley, T O., Thomas, E. 01 November 1994 (has links)
No description available.
24

Effect of Enhancement on Convolutional Neural Network Based Multi-view Object Classification

Xie, Zhiyuan 29 May 2018 (has links)
No description available.
25

Use of Multiple Imaging Views for Improving Image Quality in Small Animal MR Imaging Studies

Manivannan, Niranchana 13 October 2015 (has links)
No description available.
26

An empirical study of stability and variance reduction in DeepReinforcement Learning

Lindström, Alexander January 2024 (has links)
Reinforcement Learning (RL) is a branch of AI that deals with solving complex sequential decision making problems such as training robots, trading while following patterns and trends, optimal control of industrial processes, and more. These applications span various fields, including data science, factories, finance, and others[1]. The most popular RL algorithm today is Deep Q Learning (DQL), developed by a team at DeepMind, which successfully combines RL with Neural Network (NN). However, combining RL and NN introduces challenges such as numerical instability and unstable learning due to high variance. Among others, these issues are due to the“moving target problem”. To mitigate this problem, the target network was introduced as a solution. However, using a target network slows down learning, vastly increases memory requirements, and adds overheads in running the code. In this thesis, we conduct an empirical study to investigate the importance of target networks. We conduct this empirical study for three scenarios. In the first scenario, we train agents in online learning. The aim here is to demonstrate that the target network can be removed after some point in time without negatively affecting performance. To evaluate this scenario, we introduce the concept of the stabilization point. In thesecond scenario, we pre-train agents before continuing to train them in online learning. For this scenario, we demonstrate the redundancy of the target network by showing that it can be completely omitted. In the third scenario, we evaluate a newly developed activation function called Truncated Gaussian Error Linear Unit (TGeLU). For thisscenario, we train an agent in online learning and show that by using TGeLU as anactivation function, we can completely remove the target network. Through the empirical study of these scenarios, we conjecture and verify that a target network has only transient benefits concerning stability. We show that it has no influence on the quality of the policy found. We also observed that variance was generally higher when using a target network in the later stages of training compared to cases where the target network had been removed. Additionally, during the investigation of the second scenario, we observed that the magnitude of training iterations during pre-training affected the agent’s performance in the online learning phase. This thesis provides a deeper understanding of how the target networkaffects the training process of DQL, some of them - surrounding variance reduction- are contrary to popular belief. Additionally, the results have provided insights into potential future work. These include further explore the benefits of lower variance observed when removing the target network and conducting more efficient convergence analyses for the pre-training part in the second scenario.

Page generated in 0.2512 seconds