逆強化学習を用いたカイコガの匂い源探索における報酬関数および行動戦略の推定

抄録

In this paper, we estimated the reward function of odor source searching of bombyx mori by inverse reinforcement learning. Based on the reward function, we reconstructed the policy of odor source searching by reinforcement learning. We discussed these results in light of previous studies. We compared the policy with bombyx mori odor source searching using simulation.