arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2003.05325cs.LGstat.ML
收藏

元学习好奇算法

Meta-learning curiosity algorithms

Computer Science and Artificial Intelligence Laboratory(计算机科学与人工智能实验室);Massachusetts Institute of Technology(麻省理工学院)

Ferran Alet, Martin F. Schneider, Tomas Lozano-Perez, Leslie Pack Kaelbling

摘要:

我们假设好奇心是进化发现的一种机制,它鼓励智能体在生命早期进行有意义的探索,以使其接触到使其能够在一生中获得高回报的经验。我们将产生好奇行为的问题制定为元学习之一:外循环将搜索好奇机制的空间,这些机制动态地适应代理的奖励信号,而内循环将使用适应的奖励信号执行标准的强化学习。然而,目前基于转移神经网络权重的元强化学习方法仅在非常相似的任务之间推广。为了扩大泛化,我们提出了元学习算法:类似于人类在ML论文中设计的代码片段。我们丰富的程序语言将神经网络与其他构建模块(如缓冲区、最近邻模块和自定义损失函数)相结合。我们证明了经验的方法的有效性,找到两个新的好奇心算法,执行标准或更好地比人类设计的已发表的好奇心算法在域不同的网格导航与图像输入,杂技演员,月球着陆器,蚂蚁和料斗。

英文摘要:

We hypothesize that curiosity is a mechanism found by evolution that encourages meaningful exploration early in an agent's life in order to expose it to experiences that enable it to obtain high rewards over the course of its lifetime. We formulate the problem of generating curious behavior as one of meta-learning: an outer loop will search over a space of curiosity mechanisms that dynamically adapt the agent's reward signal, and an inner loop will perform standard reinforcement learning using the adapted reward signal. However, current meta-RL methods based on transferring neural network weights have only generalized between very similar tasks. To broaden the generalization, we instead propose to meta-learn algorithms: pieces of code similar to those designed by humans in ML papers. Our rich language of programs combines neural networks with other building blocks such as buffers, nearest-neighbor modules and custom loss functions. We demonstrate the effectiveness of the approach empirically, finding two novel curiosity algorithms that perform on par or better than human-designed published curiosity algorithms in domains as disparate as grid navigation with image inputs, acrobot, lunar lander, ant and hopper.

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。

↑