Automated Curriculum Learning for Embodied Agents: A Neuroevolutionary Approach
This work addresses the challenge of automated curriculum learning for embodied agents, offering a domain-agnostic method that is incremental in improving evolutionary algorithms.
The paper tackles the problem of automatically adjusting environmental difficulty for evolving agents in curriculum learning, demonstrating that the proposed neuroevolutionary approach outperforms conventional algorithms and generates robust solutions across varying conditions.
We demonstrate how an evolutionary algorithm can be extended with a curriculum learning process that selects automatically the environmental conditions in which the evolving agents are evaluated. The environmental conditions are selected so to adjust the level of difficulty to the ability level of the current evolving agents and so to challenge the weaknesses of the evolving agents. The method does not require domain knowledge and does not introduce additional hyperparameters. The results collected on two benchmark problems, that require to solve a task in significantly varying environmental conditions, demonstrate that the method proposed outperforms conventional algorithms and generates solutions that are robust to variations