相対的行動クレジット学習による局所解脱出
Learning from Relative Action Credit for Escaping Local Minima
PPOに行動条件付き価値評価を追加し、同一状態での代替行動と比較することで、接触の多い長期的タスクにおける局所解の停滞を脱出する手法を提案した。
詳しい要約
1. どんなもの?
2. 先行研究と比べてどこがすごい?
3. 技術・手法の肝は?
4. どうやって有効だと検証した?
5. 議論はある?
6. 次に読むべき論文は?
※ AIが要旨から生成した要約です。正確性は原文をご確認ください。
著者: Giovanni Minelli, Giulio Turrisi, Lorenzo Amatucci, Victor Barasuol, Claudio Semini
分類: cs.LG, cs.RO
原文アブストラクト
Reinforcement learning holds the promise of robots that learn manipulation skills from interaction alone, without pretrained priors or demonstration data. In practice, on-policy methods such as PPO routinely stall on long-horizon, contact-rich tasks: dense reward shaping carries the policy through stages, yet stitching them together and discovering sparsely rewarded events rarely succeeds. We trace this failure to the missing action discrimination in state-value credit assignment: a state-value critic judges progress only by episode outcomes, so when the policy converges to a partial solution all outcomes coincide, and improvement stops because no action can prove better than another. We therefore extend PPO with an action conditioned value critic which, can compare the executed action with alternatives sampled at the same state and aid the learning signal. Combined with Generalized Advantage Estimation and a bound on exploration noise, we show how the extended algorithm escapes the plateau and completes precise peg insertion within a training budget where PPO with identical exploration fails.
