The Experts below are selected from a list of 78771 Experts worldwide ranked by ideXlab platform

Fabio Ramos - One of the best experts on this subject based on the ideXlab platform.

  • Sequential Bayesian Optimisation as a POMDP for Environment Monitoring with UAVs
    2017
    Co-Authors: Philippe Morere, Roman Marchant, Fabio Ramos
    Abstract:

    Bayesian Optimisation has gained much popularity lately, as a global optimisation technique for functions that are expensive to evaluate or unknown a priori. While classical BO focuses on where to gather an observation next, it does not take into account practical constraints for a robotic system such as where it is physically possible to gather samples from, nor the sequential nature of the problem while executing a trajectory. In field robotics and other real-life situations, physical and trajectory constraints are inherent problems. This paper addresses these issues by formulating Bayesian Optimisation for continuous trajectories within a Partially Observable Markov Decision Process (POMDP) framework. The resulting POMDP is solved using Monte-Carlo Tree Search (MCTS), which we adapt to using a reward function balancing exploration and exploitation. Experiments on monitoring a Spatial Phenomenon with a UAV illustrate how our BO-POMDP algorithm outperforms competing techniques.

  • ICRA - Sequential Bayesian optimization as a POMDP for environment monitoring with UAVs
    2017 IEEE International Conference on Robotics and Automation (ICRA), 2017
    Co-Authors: Philippe Morere, Roman Marchant, Fabio Ramos
    Abstract:

    Bayesian Optimization has gained much popularity lately, as a global optimization technique for functions that are expensive to evaluate or unknown a priori. While classical BO focuses on where to gather an observation next, it does not take into account practical constraints for a robotic system such as where it is physically possible to gather samples from, nor the sequential nature of the problem while executing a trajectory. In field robotics and other real-life situations, physical and trajectory constraints are inherent problems. This paper addresses these issues by formulating Bayesian Optimization for continuous trajectories within a Partially observable Markov Decision Process (POMDP) framework. The resulting POMDP is solved using Monte-Carlo Tree Search (MCTS), which we adapt to using a reward function balancing exploration and exploitation. Experiments on monitoring a Spatial Phenomenon with a UAV illustrate how our BO-POMDP algorithm outperforms competing techniques.

Chisoni Mumba - One of the best experts on this subject based on the ideXlab platform.

  • application of system dynamics and participatory Spatial group model building in animal health a case study of east coast fever interventions in lundazi and monze districts of zambia
    PLOS ONE, 2017
    Co-Authors: Chisoni Mumba, Eystein Skjerve, Magda Rich, Karl M Rich
    Abstract:

    East Coast Fever (ECF) is the most economically important production disease among traditional beef cattle farmers in Zambia. Despite the disease control efforts by the government, donors, and farmers, ECF cases are increasing. Why does ECF oscillate over time? Can alternative approaches such as systems thinking contribute solutions to the complex ECF problem, avoid unintended consequences, and achieve sustainable results? To answer these research questions and inform the design and implementation of ECF interventions, we qualitatively investigated the influence of dynamic socio-economic, cultural, and ecological factors. We used system dynamics modelling to specify these dynamics qualitatively, and an innovative participatory framework called Spatial group model building (SGMB). SGMB uses participatory geographical information system (GIS) concepts and techniques to capture the role of Spatial Phenomenon in the context of complex systems, allowing stakeholders to identify Spatial Phenomenon directly on physical maps and integrate such information in model development. Our SGMB process convened focus groups of beef value chain stakeholders in two distinct production systems. The focus groups helped to jointly construct a series of interrelated system dynamics models that described ECF in a broader systems context. Thus, a complementary objective of this study was to demonstrate the applicability of system dynamics modelling and SGMB in animal health. The SGMB process revealed policy leverage points in the beef cattle value chain that could be targeted to improve ECF control. For example, policies that develop sustainable and stable cattle markets and improve household income availability may have positive feedback effects on investment in animal health. The results obtained from a SGMB process also demonstrated that a “one-size-fits-all” approach may not be equally effective in policing ECF in different agro-ecological zones due to the complex interactions of socio-ecological context with important, and often ignored, Spatial patterns.

Karl M Rich - One of the best experts on this subject based on the ideXlab platform.

  • application of system dynamics and participatory Spatial group model building in animal health a case study of east coast fever interventions in lundazi and monze districts of zambia
    PLOS ONE, 2017
    Co-Authors: Chisoni Mumba, Eystein Skjerve, Magda Rich, Karl M Rich
    Abstract:

    East Coast Fever (ECF) is the most economically important production disease among traditional beef cattle farmers in Zambia. Despite the disease control efforts by the government, donors, and farmers, ECF cases are increasing. Why does ECF oscillate over time? Can alternative approaches such as systems thinking contribute solutions to the complex ECF problem, avoid unintended consequences, and achieve sustainable results? To answer these research questions and inform the design and implementation of ECF interventions, we qualitatively investigated the influence of dynamic socio-economic, cultural, and ecological factors. We used system dynamics modelling to specify these dynamics qualitatively, and an innovative participatory framework called Spatial group model building (SGMB). SGMB uses participatory geographical information system (GIS) concepts and techniques to capture the role of Spatial Phenomenon in the context of complex systems, allowing stakeholders to identify Spatial Phenomenon directly on physical maps and integrate such information in model development. Our SGMB process convened focus groups of beef value chain stakeholders in two distinct production systems. The focus groups helped to jointly construct a series of interrelated system dynamics models that described ECF in a broader systems context. Thus, a complementary objective of this study was to demonstrate the applicability of system dynamics modelling and SGMB in animal health. The SGMB process revealed policy leverage points in the beef cattle value chain that could be targeted to improve ECF control. For example, policies that develop sustainable and stable cattle markets and improve household income availability may have positive feedback effects on investment in animal health. The results obtained from a SGMB process also demonstrated that a “one-size-fits-all” approach may not be equally effective in policing ECF in different agro-ecological zones due to the complex interactions of socio-ecological context with important, and often ignored, Spatial patterns.

Philippe Morere - One of the best experts on this subject based on the ideXlab platform.

  • Sequential Bayesian Optimisation as a POMDP for Environment Monitoring with UAVs
    2017
    Co-Authors: Philippe Morere, Roman Marchant, Fabio Ramos
    Abstract:

    Bayesian Optimisation has gained much popularity lately, as a global optimisation technique for functions that are expensive to evaluate or unknown a priori. While classical BO focuses on where to gather an observation next, it does not take into account practical constraints for a robotic system such as where it is physically possible to gather samples from, nor the sequential nature of the problem while executing a trajectory. In field robotics and other real-life situations, physical and trajectory constraints are inherent problems. This paper addresses these issues by formulating Bayesian Optimisation for continuous trajectories within a Partially Observable Markov Decision Process (POMDP) framework. The resulting POMDP is solved using Monte-Carlo Tree Search (MCTS), which we adapt to using a reward function balancing exploration and exploitation. Experiments on monitoring a Spatial Phenomenon with a UAV illustrate how our BO-POMDP algorithm outperforms competing techniques.

  • ICRA - Sequential Bayesian optimization as a POMDP for environment monitoring with UAVs
    2017 IEEE International Conference on Robotics and Automation (ICRA), 2017
    Co-Authors: Philippe Morere, Roman Marchant, Fabio Ramos
    Abstract:

    Bayesian Optimization has gained much popularity lately, as a global optimization technique for functions that are expensive to evaluate or unknown a priori. While classical BO focuses on where to gather an observation next, it does not take into account practical constraints for a robotic system such as where it is physically possible to gather samples from, nor the sequential nature of the problem while executing a trajectory. In field robotics and other real-life situations, physical and trajectory constraints are inherent problems. This paper addresses these issues by formulating Bayesian Optimization for continuous trajectories within a Partially observable Markov Decision Process (POMDP) framework. The resulting POMDP is solved using Monte-Carlo Tree Search (MCTS), which we adapt to using a reward function balancing exploration and exploitation. Experiments on monitoring a Spatial Phenomenon with a UAV illustrate how our BO-POMDP algorithm outperforms competing techniques.

Wenjing Lou - One of the best experts on this subject based on the ideXlab platform.

  • WSN09-3: Fault-tolerant Event Boundary Detection in Wireless Sensor Networks
    IEEE Globecom 2006, 2006
    Co-Authors: Kui Ren, Kai Zeng, Wenjing Lou
    Abstract:

    Event boundary detection is in and of itself a useful application in wireless sensor networks (WSNs). Typically, it includes the detection of a large-scale Spatial Phenomenon such as the transportation front line of a contamination or the diagnosis of network health. In this paper, we present FEBD, a fully distributed and light-weight fault-tolerant event boundary detection scheme. FEBD features an enhanced (nonparametric) statistical model that supports localized detection among neighboring nodes. To enhance detection accuracy, FEBD also introduces an error suppression technique prior to the determination of boundary nodes. The proposed scheme shows a much better detection accuracy and fault tolerance properties as compared to the previous models. The proposed FEBD is evaluated by extensive simulations, and presents very good detection accuracy, even when sensor fault probability is as high as 20%.

  • GLOBECOM - Fault-tolerant Event Boundary Detection in Wireless Sensor Networks
    2006
    Co-Authors: Kui Ren, Kai Zeng, Wenjing Lou
    Abstract:

    Event boundary detection is in and of itself a useful application in wireless sensor networks (WSNs). Typically, it includes the detection of a large-scale Spatial Phenomenon such as the transportation front line of a contamination or the diagnosis of network health. In this paper, we present FEBD, a fully distributed and light-weight Fault-tolerant Event Boundary Detection scheme. FEBD features an enhanced (nonparametric) statistical model that supports localized detection among neighboring nodes. To enhance detection accuracy, FEBD also introduces an error suppression technique prior to the determination of boundary nodes. The proposed scheme shows a much better detection accuracy and fault tolerance properties as compared to the previous models. The proposed FEBD is evaluated by extensive simulations, and presents very good detection accuracy, even when sensor fault probability is as high as 20%.