arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2211.03352cs.LG
收藏

基于课程的非对称多任务强化学习

Curriculum-based Asymmetric Multi-task Reinforcement Learning

Nanyang Technological University(南洋理工大学);Tencent Inc.(腾讯公司);JD.com Inc.(京东集团)

Hanchi Huang, Deheng Ye, Li Shen, Wei Liu

摘要:

我们介绍CAMRL,第一个基于非对称多任务学习(AMTL)算法,用于处理多个强化学习(RL)任务。为了减轻基于CAMRL的AMTL中定制一次性训练顺序的负面影响,CAMRL根据训练时间、总体性能和任务间性能差距的指标,在并行单任务RL和非对称多任务RL(MTRL)之间切换其训练模式。为了灵活地利用多源先验知识并减少AMTL中的负传递,我们定制了具有多个可微排序函数的复合损失,并通过交替优化和Frank-Wolfe算法来优化损失。基于不确定性的超参数自动调整也被应用于消除在优化过程中费力的超参数分析的需要。通过优化复合损失,CAMRL预测下一个训练任务,并不断重新访问传递矩阵和网络权重。我们已经在多任务RL中进行了广泛的基准测试,包括Gym-minigrid,Meta-world,Atari视频游戏,基于视觉的PyBullet任务和RLBench,以显示CAMRL相对于相应的单任务RL算法和最先进的MTRL算法的改进。代码可在以下网址获得:https://github.com/huanghanchi/CAMRL

英文摘要:

We introduce CAMRL, the first curriculum-based asymmetric multi-task learning (AMTL) algorithm for dealing with multiple reinforcement learning (RL) tasks altogether. To mitigate the negative influence of customizing the one-off training order in curriculum-based AMTL, CAMRL switches its training mode between parallel single-task RL and asymmetric multi-task RL (MTRL), according to an indicator regarding the training time, the overall performance, and the performance gap among tasks. To leverage the multi-sourced prior knowledge flexibly and to reduce negative transfer in AMTL, we customize a composite loss with multiple differentiable ranking functions and optimize the loss through alternating optimization and the Frank-Wolfe algorithm. The uncertainty-based automatic adjustment of hyper-parameters is also applied to eliminate the need of laborious hyper-parameter analysis during optimization. By optimizing the composite loss, CAMRL predicts the next training task and continuously revisits the transfer matrix and network weights. We have conducted experiments on a wide range of benchmarks in multi-task RL, covering Gym-minigrid, Meta-world, Atari video games, vision-based PyBullet tasks, and RLBench, to show the improvements of CAMRL over the corresponding single-task RL algorithm and state-of-the-art MTRL algorithms. The code is available at: https://github.com/huanghanchi/CAMRL

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑