Interpreting Neural-Network Players for Game 2048
Interpreting Neural-Network Players for Game 2048
复制标题
解释《2048 游戏》的神经网络玩家
DOI：
10.1109/taai.2018.00038
复制
发表时间：
2018
期刊：
影响因子：
--
通讯作者：
Madoka Teramura
中科院分区：
文献类型：
--
作者：
Kiminori Matsuzaki;Madoka Teramura
2048游戏是一款随机单人游戏，针对2048开发强大的计算机玩家是基于通过强化学习训练的N元组网络。在我们之前的研究中，我们基于卷积神经网络（CNN）开发了2048游戏的计算机玩家，并且通过实验表明，具有三个或更多卷积层的网络比具有两个卷积层的网络性能要好得多。在这项研究中，我们分析了我们的卷积神经网络的内部工作原理（即白盒方法）以确定性能差异的原因。我们的分析包括对第一层滤波器的可视化以及针对某些特定游戏状态的网络反向追踪。我们报告了关于我们用于2048游戏的卷积神经网络内部工作的一些发现。
Game 2048 is a stochastic single-player game and development of strong computer players for 2048 has been based on N-tuple networks trained by reinforcement learning. In our previous study, we developed computer players for game 2048 based on convolutional neural networks (CNNs), and showed by experiments that networks with three or more convolution layers performed much better than that with two convolution layers. In this study, we analyze the inner working of our CNNs (i.e. white box approach) to identify the reasons of the performance. Our analyses include visualization of filters in the first layers and backward trace of the networks for some specific game states. We report several findings about inner working of our CNNs for game 2048.