Poster
Architectural Complexity Measures of Recurrent Neural Networks
Saizheng Zhang · Yuhuai Wu · Tong Che · Zhouhan Lin · Roland Memisevic · Russ Salakhutdinov · Yoshua Bengio

Wed Dec 7th 06:00 -- 09:30 PM @ Area 5+6+7+8 #75 #None
in Wednesday Posters »

In this paper, we systematically analyze the connecting architectures of recurrent neural networks (RNNs). Our main contribution is twofold: first, we present a rigorous graph-theoretic framework describing the connecting architectures of RNNs in general. Second, we propose three architecture complexity measures of RNNs: (a) the recurrent depth, which captures the RNN’s over-time nonlinear complexity, (b) the feedforward depth, which captures the local input-output nonlinearity (similar to the “depth” in feedforward neural networks (FNNs)), and (c) the recurrent skip coefficient which captures how rapidly the information propagates over time. We rigorously prove each measure’s existence and computability. Our experimental results show that RNNs might benefit from larger recurrent depth and feedforward depth. We further demonstrate that increasing recurrent skip coefficient offers performance boosts on long term dependency problems.

Keywords: Deep Learning or Neural Networks

Author Information

Saizheng Zhang (University of Montreal)
Yuhuai Wu (University of Toronto)
Tong Che (IHES)
Zhouhan Lin (University of Montreal)
Roland Memisevic (University of Montreal)
Russ Salakhutdinov (University of Toronto)
Yoshua Bengio (U. Montreal)

More from the Same Authors