The interpretability of deep learning models has raised extended attention these years. It will be beneficial if we can learn an interpretable structure from deep learning models. In this paper, we focus on Recurrent Neural Networks~(RNNs) especially gated RNNs whose inner mechanism is still not clearly understood. We find that Finite State Automaton~(FSA) that processes sequential data has more interpretable inner mechanism according to the definition of interpretability and can be learned from RNNs as the interpretable structure. We propose two methods to learn FSA from RNN based on two different clustering methods. With the learned FSA and via experiments on artificial and real datasets, we find that FSA is more trustable than the RNN from which it learned, which gives FSA a chance to substitute RNNs in applications involving humans' lives or dangerous facilities. Besides, we analyze how the number of gates affects the performance of RNN. Our result suggests that gate in RNN is important but the less the better, which could be a guidance to design other RNNs. Finally, we observe that the FSA learned from RNN gives semantic aggregated states and its transition graph shows us a very interesting vision of how RNNs intrinsically handle text classification tasks.
翻译:深层次学习模式的解释性近年来引起了人们的广泛关注。 如果我们能够从深深深学习模式中学习一个可解释的结构, 将是有益的。 在本文中, 我们聚焦于经常的神经网络- (RNN) 特别是内部机制仍然不为人们所理解的封闭式网络。 我们发现, Finite State Autamaton- (FSA) 处理连续的数据根据可解释性的定义而具有更易解释的内在机制, 并且可以从区域网络网络中学习, 作为可解释性结构。 我们建议了两种方法, 以两种不同的集群方法从区域网络中学习FSA 学习FSA 。 在所学的FSA 和 人工和真实数据集实验中, 我们发现FSA 比它所学的RNN(RNN) 更可信, 这使得FA 有机会在涉及人类生活或危险设施的应用中替代 RNN(RNN) 。 此外, 我们分析了门户的数量如何影响区域网络的绩效。 我们的门很重要, 但更好些, 用来指导设计其他区域网络网络。 最后, 我们观察到, 从区域网络所学的FA所学的Smantical ligidustrations gidudustrations sh sh sh shing shing shing shing shing shing shutdds shuts shuts ex ex ex shutdddddds shuts ex shututus ex shuts shutdowdddddates shutdddddddddddddddddddddddddddddddddds.