Agostini, Alejandro Gabriel
Total activity: 10
Research group
ROBiri - IRI Robotics Group
Institute
Institute of Robotics and Industrial Informatics
E-mail
alejandro.gabriel.agostiniestudiant.upc.edu
Contact details
UPC directory Open in new window

Graphic summary
  • Show / hide key
  • Information


Scientific and technological production
  •  

1 to 10 of 10 results
  • Access to the full text
    Integrating task planning and interactive learning for robots to work in human environments  Open access

     Agostini, Alejandro Gabriel; Torras, Carme; Wörgötter, Florentin
    International Joint Conference on Artificial Intelligence
    Presentation's date: 2011
    Presentation of work at congresses

    Read the abstract Read the abstract Access to the full text Access to the full text Open in new window  Share Reference managers Reference managers Open in new window

    Human environments are challenging for robots, which need to be trainable by lay people and learn new behaviours rapidly without disrupting much the ongoing activity. A system that integrates AI techniques for planning and learning is here proposed to satisfy these strong demands. The approach rapidly learns planning operators from few action experiences using a competitive strategy where many alternatives of cause-effect explanations are evaluated in parallel, and the most successful ones are used to generate the operators. The success of a cause-effect explanation is evaluated by a probabilistic estimate that compensates the lack of experience, producing more confident estimations and speeding up the learning in relation to other known estimates. The system operates without task interruption by integrating in the planning-learning loop a human teacher that supports the planner in making decisions. All the mechanisms are integrated and synchronized in the robot using a general decision-making framework. The feasibility and scalability of the architecture are evaluated in two different robot platforms: a Stäubli arm, and the humanoid ARMAR III.

    Postprint (author’s final draft)

  • Access to the full text
    A competitive strategy for function approximation in Q-learning  Open access

     Agostini, Alejandro Gabriel; Celaya Llover, Enric
    International Joint Conference on Artificial Intelligence
    Presentation's date: 2011
    Presentation of work at congresses

    Read the abstract Read the abstract Access to the full text Access to the full text Open in new window  Share Reference managers Reference managers Open in new window

    In this work we propose an approach for generalization in continuous domain Reinforcement Learning that, instead of using a single function approximator, tries many different function approximators in parallel, each one defined in a different region of the domain. Associated with each approximator is a relevance function that locally quantifies the quality of its approximation, so that, at each input point, the approximator with highest relevance can be selected. The relevance function is defined using parametric estimations of the variance of the q-values and the density of samples in the input space, which are used to quantify the accuracy and the confidence in the approximation, respectively. These parametric estimations are obtained from a probability density distribution represented as a Gaussian Mixture Model embedded in the input-output space of each approximator. In our experiments, the proposed approach required a lesser number of experiences for learning and produced more stable convergence profiles than when using a single function approximator.

  • Object-action complexes: grounded abstractions of sensory-motor processes

     Krüger, Norbert; Geib, Cristopher; Piater, Justus; Petrick, Ronald; Steedman, Mark; Wörgötter, Florentin; Ude, Ales; Asfour, Tamim; Kraft, Dirk; Omrcen, Damir; Agostini, Alejandro Gabriel; Dillmann, Rudiger
    Robotics and autonomous systems
    Date of publication: 2011
    Journal article

    View View Open in new window  Share Reference managers Reference managers Open in new window

  • Access to the full text
    A general strategy for interactive decision-making in robotic platforms  Open access

     Agostini, Alejandro Gabriel; Torras, Carme; Wörgötter, Florentin
    Date: 2011
    Report

    Read the abstract Read the abstract Access to the full text Access to the full text Open in new window  Share Reference managers Reference managers Open in new window

    This work presents an intergated strategy for planning and learning suitable to execute tasks with robotic platforms without any previous task specification. The approach rapidly learns planning operators from few action experiences using a competitive strategy where many alternatives of cause-effect explanations are evaluated in parallel, and the most successful ones are used to generate the operators. The system operates without task interruption by integrating in the planning-learning loop a human teacher that supports the planner in making decisions. All the mechanisms are integrated and synchronized in the robot using a general decision-making framework.

  • Access to the full text
    Stochastic approximations of average values using proportions of samples  Open access

     Agostini, Alejandro Gabriel; Celaya Llover, Enric
    Date: 2011
    Report

    Read the abstract Read the abstract Access to the full text Access to the full text Open in new window  Share Reference managers Reference managers Open in new window

    In this work we explain how the stochastic approximation of the average of a random variable is carried out when the observations used in the updates consist in proportion of samples rather than complete samples.

  • Reinforcement learning for robot control using probability density estimations

     Agostini, Alejandro Gabriel; Celaya Llover, Enric
    International Conference on Informatics in Control, Automation and Robotics
    Presentation's date: 2010
    Presentation of work at congresses

    View View Open in new window  Share Reference managers Reference managers Open in new window

  • Access to the full text
    Reinforcement learning with a Gaussian mixture model  Open access

     Agostini, Alejandro Gabriel; Celaya Llover, Enric
    International Conference on Neural Networks
    Presentation's date: 2010
    Presentation of work at congresses

    Read the abstract Read the abstract Access to the full text Access to the full text Open in new window  Share Reference managers Reference managers Open in new window

    Recent approaches to Reinforcement Learning (RL) with function approximation include Neural Fitted Q Iteration and the use of Gaussian Processes. They belong to the class of fitted value iteration algorithms, which use a set of support points to fit the value-function in a batch iterative process. These techniques make efficient use of a reduced number of samples by reusing them as needed, and are appropriate for applications where the cost of experiencing a new sample is higher than storing and reusing it, but this is at the expense of increasing the computational effort, since these algorithms are not incremental. On the other hand, non-parametric models for function approximation, like Gaussian Processes, are preferred against parametric ones, due to their greater flexibility. A further advantage of using Gaussian Processes for function approximation is that they allow to quantify the uncertainty of the estimation at each point. In this paper, we propose a new approach for RL in continuous domains based on Probability Density Estimations. Our method combines the best features of the previous methods: it is non-parametric and provides an estimation of the variance of the approximated function at any point of the domain. In addition, our method is simple, incremental, and computationally efficient. All these features make this approach more appealing than Gaussian Processes and fitted value iteration algorithms in general.

  • Quick learning of cause-effects relevant for robot action

     Agostini, Alejandro Gabriel; Wörgötter, Florentin; Torras, Carme
    Date: 2010
    Report

     Share Reference managers Reference managers Open in new window

  • Access to the full text
    Cognitive agents: a procedural perspective relying on the predictability of object-action-omplexes (OACs)  Open access

     Wörgötter, Florentin; Agostini, Alejandro Gabriel; Krüger, Norbert; Shylo, Natalya; Porr, Ben
    Robotics and autonomous systems
    Date of publication: 2009
    Journal article

    Read the abstract Read the abstract Access to the full text Access to the full text Open in new window  Share Reference managers Reference managers Open in new window

    Embodied cognition suggests that complex cognitive traits can only arise when agents have a body situated in the world. The aspects of embodiment and situatedness are being discussed here from the perspective of linear systems theory. This perspective treats bodies as dynamic, temporally variable entities, which can be extended (or curtailed) at their boundaries. We show how acting agents can, for example, actively extend their body for some time by incorporating predictably behaving parts of the world and how this affects the transfer functions. We suggest that primates have mastered this to a large degree increasingly splitting their world into predictable and unpredictable entities. We argue that temporary body extension may have been instrumental in paving the route for the development higher cognitive complexity as it is reliably widening the cause-effect horizon about the actions of the agent. A first robot experiment is sketched to support these ideas. We continue discussing the concept of Object-Action Complexes (OACs) introduced by the European PACO-PLUS consortium to emphasize the notion that for a cognitive agent objects and actions are inseparably intertwined. In another robot experiment we devise a semi-supervised procedure using the OAC-concept to demonstrate how an agent can acquire knowledge about its world. Here the notion of predicting changes fundamentally underlies the implemented procedure and we try to show how this concept can be used to improve the robot’s inner model and behaviour. Hence, in this article we have tried to show how predictability can be used to augment the agent’s body and to acquire knowledge about the external world, possibly leading to more advanced cognitive traits.

    Postprint (author’s final draft)

  • Action rule induction from cause-effect pairs learned through robot-teacher interaction

     Agostini, Alejandro Gabriel; Celaya Llover, Enric; Torras, Carme; Wörgötter, Florentin
    International Conference on Cognitive Systems
    Presentation's date: 2008
    Presentation of work at congresses

    View View Open in new window  Share Reference managers Reference managers Open in new window