Exploration strategies for homeostatic agents

Paper in proceeding, 2019

This paper evaluates two new strategies for investigating artificial animals called animats. Animats are homeostatic agents with the objective of keeping their internal variables as close to optimal as possible. Steps towards the optimal are rewarded and steps away punished. Using reinforcement learning for exploration and decision making, the animats can consider predetermined optimal/acceptable levels in light of current levels, giving them greater flexibility for exploration and better survival chances. This paper considers the resulting strategies as evaluated in a range of environments, showing them to outperform common reinforcement learning, where internal variables are not taken into consideration.