Efficient learning of reactive robot behaviors with a Neural-Q_learning approach

The purpose of this paper is to propose a Neural-Q_learning approach designed for online learning of simple and reactive robot behaviors. In this approach, the Q_function is generalized by a multi-layer neural network allowing the use of continuous states and actions. The algorithm uses a database of the most recent learning samples to accelerate and guarantee the convergence. Each Neural-Q_learning function represents an independent, reactive and adaptive behavior which maps sensorial states to robot control actions. A group of these behaviors constitutes a reactive control scheme designed to fulfill simple missions. The paper centers on the description of the Neural-Q_learning based behaviors showing their performance with an underwater robot in a target following task. Real experiments demonstrate the convergence and stability of the learning system, pointing out its suitability for online robot learning. Advantages and limitations are discussed

Testing Delivery Systems in Transnational virtual learning: the vocational management training for the european tourism industry (VocMat) case study

This article discusses the lessons learned from developing and delivering the Vocational Management Training for the European Tourism Industry (VocMat) online training programme, which was aimed at providing flexible, online distance learning for the European tourism industry. The programme was designed to address managers ‘need for flexible, senior management level training which they could access at a time and place which fitted in with their working and non-work commitments. The authors present two main approaches to using the Virtual Learning Environment, the feedback from the participants, and the implications of online Technology in extending tourism training opportunities

Entorno de aprendizaje virtual adaptativo soportado por un modelo de usuario integral

El objetivo de esta tesis es mejorar la efectividad y eficiencia de los entornos de aprendizaje virtual. Para lograr este propósito se define un Modelo de Usuario que considera las características del usuario, el contexto y la Interacción. Estas tres dimensiones son integradas en un Modelo de Usuario Integral (MUI) para proveer adaptación de contenido, formato y actividades en entornos educativos con heterogeneidad de usuarios, tecnologías e interacciones. Esta heterogeneidad genera la entrega de contenidos, formatos y actividades inadecuadas para los estudiantes. La particularización del MUI en un entorno educativo es definida Modelo de Estudiante Integral (MEI). Las principales aportaciones de esta tesis son la definición y validación de un MUI, la utilización de un MEI abierto para propiciar la reflexión de los estudiantes sobre sus procesos de aprendizaje, la integración tecnológica con independencia de plataforma y la validación del MEI con estudiantes en escenarios reales.

Semi-online neural-Q_leaming for real-time robot learning

Reinforcement learning (RL) is a very suitable technique for robot learning, as it can learn in unknown environments and in real-time computation. The main difficulties in adapting classic RL algorithms to robotic systems are the generalization problem and the correct observation of the Markovian state. This paper attempts to solve the generalization problem by proposing the semi-online neural-Q_learning algorithm (SONQL). The algorithm uses the classic Q_learning technique with two modifications. First, a neural network (NN) approximates the Q_function allowing the use of continuous states and actions. Second, a database of the most representative learning samples accelerates and stabilizes the convergence. The term semi-online is referred to the fact that the algorithm uses the current but also past learning samples. However, the algorithm is able to learn in real-time while the robot is interacting with the environment. The paper shows simulated results with the "mountain-car" benchmark and, also, real results with an underwater robot in a target following behavior