Online Associative Multi-Stage Goal Babbling Toward Versatile Learning of Sensorimotor Skills

Rayyes, Rania; Steil, Jochen J.

doi:10.1109/devlrn.2019.8850707

Cited by 4 publications

(3 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…When facing the problem of autonomously learning multiple tasks, researchers typically focus on solutions where just one parametrised policy per task is sufficient to solve them [24], [25], [26]. Redundancy in motor control faced with autonomous agents is usually addressed as a problem and tackled through strategies such as goal-babbling [27], [28], which has been successfully implemented together with intrinsic motivation [29]. However, in complex environments presenting different contexts in time, the same task might need a set of different skills to be solved.…”

Section: Introductionmentioning

confidence: 99%

Autonomous Reinforcement Learning of Multiple Interrelated Tasks

Santucci

Baldassarre

Cartoni

2019

2019 Joint IEEE 9th International Conference on Development and Learning and Epigenetic Robotics (ICDL-EpiRob)

View full text Add to dashboard Cite

Autonomous multiple tasks learning is a fundamental capability to develop versatile artificial agents that can act in complex environments. In real-world scenarios, tasks may be interrelated (or "hierarchical") so that a robot has to first learn to achieve some of them to set the preconditions for learning other ones. Even though different strategies have been used in robotics to tackle the acquisition of interrelated tasks, in particular within the developmental robotics framework, autonomous learning in this kind of scenarios is still an open question. Building on previous research in the framework of intrinsically motivated open-ended learning, in this work we describe how this question can be addressed working on the level of task selection, in particular considering the multiple interrelated tasks scenario as an MDP where the system is trying to maximise its competence over all the tasks.

show abstract

Section: Introductionmentioning

confidence: 99%

Autonomous Reinforcement Learning of Multiple Interrelated Tasks

Santucci

Baldassarre

Cartoni

2019

2019 Joint IEEE 9th International Conference on Development and Learning and Epigenetic Robotics (ICDL-EpiRob)

View full text Add to dashboard Cite

show abstract

“…2 The Minisom library (https://github.com/JustGlowing/minisom) has been re-adapted in the experiment presented here. 3 This diverges from the classical view of internal models. We believe however that this has a limited conceptual impact on the architecture.…”

Section: Online Learning Of Inverse and Forward Modelsmentioning

confidence: 93%

“…On the contrary, the forward model is in charge of predicting the sensory outcome of a given motor command that is passed as input. It has to be noted that, in the experiments presented here, the motor commands are absolute target motor positions, therefore no initial sensory state is needed as input in both internal models 3 .…”

Section: Online Learning Of Inverse and Forward Modelsmentioning

confidence: 99%

Tracking Emotions: Intrinsic Motivation Grounded on Multi-Level Prediction Error Dynamics

Schillaci¹,

Ciria²,

Lara³

2020

Preprint

View full text Add to dashboard Cite

How do cognitive agents decide which is the relevant information to learn and how goals are selected to gain this knowledge? Cognitive agents need to be motivated to perform any action. We discuss that emotions arise when differences between expected and actual rates of progress toward a goal are experienced. Therefore, the tracking of prediction error dynamics has a tight relationship with emotions. Here, we suggest that the tracking of prediction error dynamics allows an artificial agent to be intrinsically motivated to seek new experiences but constrained to those that generate reducible prediction error. We present an intrinsic motivation architecture that generates behaviors towards self-generated and dynamic goals and that regulates goal selection and the balance between exploitation and exploration through multi-level monitoring of prediction error dynamics. This new architecture modulates exploration noise and leverages computational resources according to the dynamics of the overall performance of the learning system. Additionally, it establishes a possible solution to the temporal dynamics of goal selection. The results of the experiments presented here suggest that this architecture outperforms intrinsic motivation approaches where exploratory noise and goals are fixed and a greedy strategy is applied.

show abstract