• Refine Query
  • Source
  • Publication year
  • to
  • Language
  • 7
  • 2
  • 2
  • 1
  • Tagged with
  • 13
  • 13
  • 8
  • 7
  • 5
  • 5
  • 3
  • 3
  • 3
  • 3
  • 3
  • 2
  • 2
  • 2
  • 2
  • 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.
1

Využití evolučních technik v hierarchickém plánování / Evolutionary techniques utilization in hierarchical task network

Řeháková, Lucie January 2016 (has links)
This master thesis describes the design and the implementation of the algorithm solving the domain- independent partial order simple task network planning problem using the tree-based genetic programming. The work contains comparison of several possible approaches to the problem --- it compares different representations, ways of evaluation and approaches to the partial ordering. It defines heuristics to improve the efficiency of the algorithm, including the distance heuristic, the local search and the individual equivalency. The implementation was tested on several experiments to show the abilities, strengths and weaknesses of the algorithm. Powered by TCPDF (www.tcpdf.org)
2

Raffinement des intentions / Refinement of Intentions

Xiao, Zhanhao 12 December 2017 (has links)
Le résumé en français n'a pas été communiqué par l'auteur. / Le résumé en anglais n'a pas été communiqué par l'auteur.
3

A Qualitative Study of Task and Work-Social Networks

Sausan, Nabeela January 2012 (has links)
Despite the well-recognized importance of interpersonal relationships within the work environment, there is no comprehensive approach or set of studies that provide a complete picture. As a step towards providing a complete picture, this research presents a qualitative exploratory study of how people experience the work environment through task and work-social interactions and through policies and norms present in the work environment. The purpose of this study is to understand the overall work experience from purely work-social and purely task network perspective. A semi-structured question-based set of interviews were performed among professionals from a Canadian university alumni society. The transcripts of the interviews were then manually coded and analyzed using statistical methods. The study found an overall higher level of positive responses for co-workers in the work-social network, as well as a preference for work-social co-workers in building a team for a hypothetical company. The study also found a general willingness to select only the best ranked co-workers from task network, whereas a leniency towards the selection of co-workers from work-social network was observed despite their lower ranking. At the same time, the study identified some of the most desirable attributes of fellow co-workers both in the task and work-social networks from an employee’s perspective. The significance of the people and team was found more important to the employees than factors such as specific task performed or compensation. This exploratory study provided insights into how employees view their co-workers and their work environment.
4

A Qualitative Study of Task and Work-Social Networks

Sausan, Nabeela January 2012 (has links)
Despite the well-recognized importance of interpersonal relationships within the work environment, there is no comprehensive approach or set of studies that provide a complete picture. As a step towards providing a complete picture, this research presents a qualitative exploratory study of how people experience the work environment through task and work-social interactions and through policies and norms present in the work environment. The purpose of this study is to understand the overall work experience from purely work-social and purely task network perspective. A semi-structured question-based set of interviews were performed among professionals from a Canadian university alumni society. The transcripts of the interviews were then manually coded and analyzed using statistical methods. The study found an overall higher level of positive responses for co-workers in the work-social network, as well as a preference for work-social co-workers in building a team for a hypothetical company. The study also found a general willingness to select only the best ranked co-workers from task network, whereas a leniency towards the selection of co-workers from work-social network was observed despite their lower ranking. At the same time, the study identified some of the most desirable attributes of fellow co-workers both in the task and work-social networks from an employee’s perspective. The significance of the people and team was found more important to the employees than factors such as specific task performed or compensation. This exploratory study provided insights into how employees view their co-workers and their work environment.
5

Artificiell intelligens : Prestanda hos beteendeträd och Hierarchical Task Network / Artificial intelligence : Performance of behavior trees and Hierarchical Task Network.

Juvél, Joel January 2014 (has links)
Detta arbete undersöker skillnader i tidseffektivitet mellan beteendeträd ochHierarchical Task Network. En enklare spelprototyp av typen top-down 2D shoot-emup implementerades. Spelprototypen använder två typer av autonom motspelare så kallade botar. En bot för beteendeträd och en bot för Hierarchical Task Network. Spelprototypen mäter körtiden för varje typ av bot i sex olika situationer. Varje situation svarar mot ett bestämt beteende hos boten. Ett beteende kan brytas ned i en samling uppgifter som boten kan utföra. Spelprototypen mäter körtiden för att bestämma ett enskilt beteende i en specifik situation. Resultaten från mätningarna tyder på att beteendeträd har bättre tidseffektivitet än Hierarchical Task Network.
6

Improving Multi-Task Learning in Autonomous Driving Perception with Dynamic Loss Weights and Individual Encoders / Förbättrande av multi-task learning i autonom körning med dynamiska viktminskningar och enskilda encoders

Jiang, Zehao January 2024 (has links)
The perception tasks in autonomous driving, namely 3D object detection and map segmentation, play a crucial role in enabling vehicles to perceive the surrounding environment. The traditional approach is to have a single network for each task and complete all tasks in a sequential manner. However, this method suffers from repeated feature extraction and error propagation, leading to inefficiency and reduced accuracy. While multi-task learning can eliminate redundant computations and facilitate information exchange among tasks, improving efficiency and overall system performance, it can also lead to a reduction in the performance of a particular task, compared to single-task training due to gradient dominance. To tackle this problem, this thesis aims to bridge the performance gap between multi-task and single-task learning. We first utilize the GradNorm method to dynamically readjust the loss weights while training. We further add individual encoders to allow fine-grained feature learning for each task. Based on the existing perception network, we adapt our dynamic loss strategy and new encoder architecture, which shows that our results match or even surpass the performance of each task in a multitask setting, compared to the single task. We also evaluate the computational efficiency of our method, further demonstrating the advantages of multi-task learning in the autonomous driving domain where real-time computing is non-negotiable. / Perception inom autonom körning, det vill säga 3D objektsdetektion och map segmentation, spelar en avgörande roll för att möjliggöra att fordon uppfattar den omgivande miljön. Traditionellt sett utförs detta av en grupp nätverk, en för varje uppgift, där uppgifterna utförs sekventiellt. Detta skapar problem i form av upprepad extrahering av egenskaper i datan och felfortplantning vilket leder till försämrad beräkningshastighet och resultat. Multi-task learning kan eliminera onödiga beräkningar och möjliggöra utbyte av information mellan uppgifter, vilket medför förbättringar inom effektivitet och systemets generella prestanda i relation till single-task learning. Däremot kan det leda till försämrade resultat i enskilda uppgifter på grund av gradient dominans. Denna avhandling syftar till att bemöta detta problem genom att överbrygga gapet mellan multi-task och single-task learning. Vi använder oss av GradNorm metoden för att dynamiskt justera gradienternas magnitud under träning. Enskilda encoders lades till i vardera nätverk för att möjliggöra fine-grained feature learning för varje uppgift. GradNorm och ytterligare encoders applicerades på det befintliga perception-nätverket, vilket gav resultat som är lika bra eller bättre vid multi-task learning som single-task learning för varje uppgift. Även effektivitet vid beräkning utvärderades, vilket ytterligare visade fördelarna av att använda multi-task learning i autonom körning där beräkning i realtid är av högsta prioritet.
7

Customizing the Composition of Web Services and Beyond

Sohrabi Araghi, Shirin 16 December 2013 (has links)
Web services provide a standardized means of publishing diverse, distributed applications. Increasingly, corporations are providing services or programs within and between organizations either on corporate intranets or on the cloud. Many of these services can be composed together, ideally automatically, to provide value-added service. Automated Web service composition is an example of such automation where given a specification of an objective to be realized and some knowledge of the state of the world, the problem is to automatically select, integrate, and invoke multiple services to achieve the specified objective. A popular approach to the Web service composition problem is to conceive it as an Artificial Intelligence planning task. This enables us to bring to bear many of the theoretical and computational advances in reasoning about actions to the task of Web service composition. However, Web service composition goes far beyond the reaches of classical planning, presenting a number of interesting challenges relevant to a large body of problems related to the composition of actions, programs, and services. Among these, an important challenge is generating not only a composition, but a high-quality composition tailored to user preferences. In this thesis, we present an approach to the Web service composition problem with a particular focus on the customization of compositions. We claim that there is a correspondence between generating a customized composition of Web services and non-classical Artificial Intelligence planning where the objective of the planning problem is specified as a form of control knowledge, such as a workflow or template, together with a set of constraints to be optimized or enforced. We further claim that techniques in (preference-based) planning can provide a computational basis for the development of effective, state-of-the-art techniques for generating customized compositions of Web services. To evaluate our claim, we characterize the Web service composition problem with customization as a non-classical planning problem, exploit and advance preference specification languages and preference-based planning, develop algorithms tailored to the Web service composition problem, prove formal properties of these algorithms, implement proof-of-concept systems, and evaluate these systems experimentally. While our research has been motivated by Web services, the theory and techniques we have developed are amenable to analogous problems in such diverse sectors as multi-agent systems, business process modeling, component software composition, and social and computational behaviour modeling and verification.
8

Customizing the Composition of Web Services and Beyond

Sohrabi Araghi, Shirin 16 December 2013 (has links)
Web services provide a standardized means of publishing diverse, distributed applications. Increasingly, corporations are providing services or programs within and between organizations either on corporate intranets or on the cloud. Many of these services can be composed together, ideally automatically, to provide value-added service. Automated Web service composition is an example of such automation where given a specification of an objective to be realized and some knowledge of the state of the world, the problem is to automatically select, integrate, and invoke multiple services to achieve the specified objective. A popular approach to the Web service composition problem is to conceive it as an Artificial Intelligence planning task. This enables us to bring to bear many of the theoretical and computational advances in reasoning about actions to the task of Web service composition. However, Web service composition goes far beyond the reaches of classical planning, presenting a number of interesting challenges relevant to a large body of problems related to the composition of actions, programs, and services. Among these, an important challenge is generating not only a composition, but a high-quality composition tailored to user preferences. In this thesis, we present an approach to the Web service composition problem with a particular focus on the customization of compositions. We claim that there is a correspondence between generating a customized composition of Web services and non-classical Artificial Intelligence planning where the objective of the planning problem is specified as a form of control knowledge, such as a workflow or template, together with a set of constraints to be optimized or enforced. We further claim that techniques in (preference-based) planning can provide a computational basis for the development of effective, state-of-the-art techniques for generating customized compositions of Web services. To evaluate our claim, we characterize the Web service composition problem with customization as a non-classical planning problem, exploit and advance preference specification languages and preference-based planning, develop algorithms tailored to the Web service composition problem, prove formal properties of these algorithms, implement proof-of-concept systems, and evaluate these systems experimentally. While our research has been motivated by Web services, the theory and techniques we have developed are amenable to analogous problems in such diverse sectors as multi-agent systems, business process modeling, component software composition, and social and computational behaviour modeling and verification.
9

The COMPASS Paradigm For The Systematic Evaluation Of U.S. Army Command And Control Systems Using Neural Network And Discrete Event Computer Simulation

Middlebrooks, Sam E. 15 April 2003 (has links)
In today's technology based society the rapid proliferation of new machines and systems that would have been undreamed of only a few short years ago has become a way of life. Developments and advances especially in the areas of digital electronics and micro-circuitry have spawned subsequent technology based improvements in transportation, communications, entertainment, automation, the armed forces, and many other areas that would not have been possible otherwise. This rapid "explosion" of new capabilities and ways of performing tasks has been motivated as often as not by the philosophy that if it is possible to make something better or work faster or be more cost effective or operate over greater distances then it must inherently be good for the human operator. Taken further, these improvements typically are envisioned to consequently produce a more efficient operating system where the human operator is an integral component. The formal concept of human-system interface design has only emerged this century as a recognized academic discipline, however, the practice of developing ideas and concepts for systems containing human operators has been in existence since humans started experiencing cognitive thought. An example of a human system interface technology for communication and dissemination of written information that has evolved over centuries of trial and error development, is the book. It is no accident that the form and shape of the book of today is as it is. This is because it is a shape and form readily usable by human physiology whose optimal configuration was determined by centuries of effort and revision. This slow evolution was mirrored by a rate of technical evolution in printing and elsewhere that allowed new advances to be experimented with as part of the overall use requirement and need for the existence of the printed word and some way to contain it. Today, however, technology is advancing at such a rapid rate that evolutionary use requirements have no chance to develop along side the fast pace of technical progress. One result of this recognition is the establishment of disciplines like human factors engineering that have stated purposes and goals of systematic determination of good and bad human system interface designs. However, other results of this phenomenon are systems that get developed and placed into public use simply because new technology allowed them to be made. This development can proceed without a full appreciation of how the system might be used and, perhaps even more significantly, what impact the use of this new system might have on the operator within it. The U.S. Army has a term for this type of activity. It is called "stove-piped development". The implication of this term is that a system gets developed in isolation where the developers are only looking "up" and not "around". They are thus concerned only with how this system may work or be used for its own singular purposes as opposed to how it might be used in the larger community of existing systems and interfaces or, even more importantly, in the larger community of other new systems in concurrent development. Some of the impacts for the Army from this mode of system development are communication systems that work exactly as designed but are unable to interface to other communications systems in other domains for battlefield wide communications capabilities. Having communications systems that cannot communicate with each other is a distinct problem in its own right. However, when developments in one industry produce products that humans use or attempt to use with products from totally separate developments or industries, the Army concept of product development resulting from stove-piped design visions can have significant implication on the operation of each system and the human operator attempting to use it. There are many examples that would illustrate the above concept, however, one that will be explored here is the Army effort to study, understand, and optimize its command and control (C2) operations. This effort is at the heart of a change in the operational paradigm in C2 Tactical Operations Centers (TOCs) that the Army is now undergoing. For the 50 years since World War II the nature, organization, and mode of the operation of command organizations within the Army has remained virtually unchanged. Staffs have been organized on a basic four section structure and TOCs generally only operate in a totally static mode with the amount of time required to move them to keep up with a mobile battlefield going up almost exponentially from lower to higher command levels. However, current initiatives are changing all that and while new vehicles and hardware systems address individual components of the command structures to improve their operations, these initiatives do not necessarily provide the environment in which the human operator component of the overall system can function in a more effective manner. This dissertation examines C2 from a system level viewpoint using a new paradigm for systematically examining the way TOCs operate and then translating those observations into validated computer simulations using a methodological framework. This paradigm is called COmputer Modeling Paradigm And Simulation of Systems (COMPASS). COMPASS provides the ability to model TOC operations in a way that not only includes the individuals, work groups and teams in it, but also all of the other hardware and software systems and subsystems and human-system interfaces that comprise it as well as the facilities and environmental conditions that surround it. Most of the current literature and research in this area focuses on the concept of C2 itself and its follow-on activities of command, control, communications (C3), command, control, communications, and computers (C4), and command, control, communications, computers and intelligence (C4I). This focus tends to address the activities involved with the human processes within the overall system such as individual and team performance and the commander's decision-making process. While the literature acknowledges the existence of the command and control system (C2S), little effort has been expended to quantify and analyze C2Ss from a systemic viewpoint. A C2S is defined as the facilities, equipment, communications, procedures, and personnel necessary to support the commander (i.e., the primary decision maker within the system) for conducting the activities of planning, directing, and controlling the battlefield within the sector of operations applicable to the system. The research in this dissertation is in two phases. The overall project incorporates sequential experimentation procedures that build on successive TOC observation events to generate an evolving data store that supports the two phases of the project. Phase I consists of the observation of heavy maneuver battalion and brigade TOCs during peacetime exercises. The term "heavy maneuver" is used to connotate main battle forces such as armored and mechanized infantry units supported by artillery, air defense, close air, engineer, and other so called combat support elements. This type of unit comprises the main battle forces on the battlefield. It is used to refer to what is called the conventional force structure. These observations are conducted using naturalistic observation techniques of the visible functioning of activities within the TOC and are augmented by automatic data collection of such things as analog and digital message traffic, combat reports generated by the computer simulations supporting the wargame exercise, and video and audio recordings where appropriate and available. Visible activities within the TOC include primarily the human operator functions such as message handling activities, decision-making processes and timing, coordination activities, and span of control over the battlefield. They also include environmental conditions, functional status of computer and communications systems, and levels of message traffic flows. These observations are further augmented by observer estimations of such indicators as perceived level of stress, excitement, and level of attention to the mission of the TOC personnel. In other words, every visible and available component of the C2S within the TOC is recorded for analysis. No a priori attempt is made to evaluate the potential significance of each of the activities as their contribution may be so subtle as to only be ascertainable through statistical analysis. Each of these performance activities becomes an independent variable (IV) within the data that is compared against dependent variables (DV) identified according to the mission functions of the TOC. The DVs for the C2S are performance measures that are critical combat tasks performed by the system. Examples of critical combat tasks are "attacking to seize an objective", "seizure of key terrain", and "river crossings'. A list of expected critical combat tasks has been prepared from the literature and subject matter expert (SME) input. After the exercise is over, the success of these critical tasks attempted by the C2S during the wargame are established through evaluator assessments, if available, and/or TOC staff self analysis and reporting as presented during after action reviews. The second part of Phase I includes datamining procedures, including neural networks, used in a constrained format to analyze the data. The term constrained means that the identification of the outputs/DV is known. The process was to identify those IV that significantly contribute to the constrained DV. A neural network is then constructed where each IV forms an input node and each DV forms an output node. One layer of hidden nodes is used to complete the network. The number of hidden nodes and layers is determined through iterative analysis of the network. The completed network is then trained to replicate the output conditions through iterative epoch executions. The network is then pruned to remove input nodes that do not contribute significantly to the output condition. Once the neural network tree is pruned through iterative executions of the neural network, the resulting branches are used to develop algorithmic descriptors of the system in the form of regression like expressions. For Phase II these algorithmic expressions are incorporated into the CoHOST discrete event computer simulation model of the C2S. The programming environment is the commercial programming language Micro Saintä running on a PC microcomputer. An interrogation approach was developed to query these algorithms within the computer simulation to determine if they allow the simulation to reflect the activities observed in the real TOC to within an acceptable degree of accuracy. The purpose of this dissertation is to introduce the COMPASS concept that is a paradigm for developing techniques and procedures to translate as much of the performance of the entire TOC system as possible to an existing computer simulation that would be suitable for analyses of future system configurations. The approach consists of the following steps: • Naturalistic observation of the real system using ethnographic techniques. • Data analysis using datamining techniques such as neural networks. • Development of mathematical models of TOC performance activities. • Integration of the mathematical into the CoHOST computer simulation. • Interrogation of the computer simulation. • Assessment of the level of accuracy of the computer simulation. • Validation of the process as a viable system simulation approach. / Ph. D.
10

A decentralised online multi-agent planning framework for multi-agent systems

Cardoso, Rafael Cau? 27 March 2018 (has links)
Submitted by PPG Ci?ncia da Computa??o (ppgcc@pucrs.br) on 2018-05-08T18:37:11Z No. of bitstreams: 1 RAFAEL_CAU?_CARDOSO_TES.pdf: 14431785 bytes, checksum: 227194ed28a9e55c3ab1fbedebf06922 (MD5) / Approved for entry into archive by Caroline Xavier (caroline.xavier@pucrs.br) on 2018-05-15T19:07:36Z (GMT) No. of bitstreams: 1 RAFAEL_CAU?_CARDOSO_TES.pdf: 14431785 bytes, checksum: 227194ed28a9e55c3ab1fbedebf06922 (MD5) / Made available in DSpace on 2018-05-15T19:14:18Z (GMT). No. of bitstreams: 1 RAFAEL_CAU?_CARDOSO_TES.pdf: 14431785 bytes, checksum: 227194ed28a9e55c3ab1fbedebf06922 (MD5) Previous issue date: 2018-03-27 / Sistemas multiagentes freq?entemente cont?m ambientes complexos e din?micos, nos quais os planos dos agentes podem falhar a qualquer momento durante a execu??o do sistema. Al?m disso, novos objetivos podem aparecer para os quais n?o existem nenhum plano dispon?vel. T?cnicas de planejamento s?o bem adequadas para lidar com esses problemas. H? uma quantidade extensa de pesquisa em planejamento centralizado para um ?nico agente, por?m, at? ent?o planejamento multiagente n?o foi completamente explorado na pr?tica. Plataformas multiagentes tipicamente proporcionam diversos mecanismos para coordena??o em tempo de execu??o, frequentemente necess?rios em planejamento online. Neste contexto, planejamento multiagente descentralizado pode ser eficiente e eficaz, especialmente em dom?nios fracamente acoplados, al?m de garantir algumas propriedades importantes em sistemas de agentes como privacidade e autonomia. N?s abordamos esse problema ao apresentar uma t?cnica para planejamento multiagente online que combina aloca??o de objetivos, planejamento individual utilizando rede de tarefas hier?rquicas (HTN), e coordena??o em tempo de execu??o para apoiar a realiza??o de objetivos sociais em sistemas multiagentes. Especificamente, n?s apresentamos um framework chamado Decentralised Online Multi-Agent Planning (DOMAP). Experimentos com tr?s dom?nios fracamente acoplados demonstram que DOMAP supera quatro planejadores multiagente do estado da arte com respeito a tempo de planejamento e tempo de execu??o, particularmente nos problemas mais dif?ceis. / Multi-agent systems often contain dynamic and complex environments where agents? course of action (plans) can fail at any moment during execution of the system. Furthermore, new goals can emerge for which there are no known plan available in any of the agents? plan library. Automated planning techniques are well suited to tackle both of these issues. Extensive research has been done in centralised planning for singleagents, however, so far multi-agent planning has not been fully explored in practice. Multi-agent platforms typically provide various mechanisms for runtime coordination, which are often required in online planning (i.e., planning during runtime). In this context, decentralised multi-agent planning can be efficient as well as effective, especially in loosely-coupled domains, besides also ensuring important properties in agent systems such as privacy and autonomy. We address this issue by putting forward an approach to online multi-agent planning that combines goal allocation, individual Hierarchical Task Network (HTN) planning, and coordination during runtime in order to support the achievement of social goals in multi-agent systems. In particular, we present a planning and execution framework called Decentralised Online Multi-Agent Planning (DOMAP). Experiments with three loosely-coupled planning domains show that DOMAP outperforms four other state-of-the-art multi agent planners with regards to both planning and execution time, particularly in the most difficult problems.

Page generated in 0.0391 seconds