arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2006.07522cs.LGstat.ML
收藏

通过信息瓶颈理解二进制神经网络的学习动力学

Understanding Learning Dynamics of Binary Neural Networks via Information Bottleneck

IIT Madras(印度理工学院马德拉斯分校)

Vishnu Raj, Nancy Nayak, Sheetal Kalyani

摘要:

紧凑的神经网络对于经济实惠且节能的深度学习解决方案至关重要。二进制神经网络(BNN)通过将权重和激活约束到两个级别($\{+1, -1\}$),将紧致化发挥到极致。然而,由于激活函数的不连续性,训练BNN并不容易,并且BNN的训练动力学还没有得到很好的理解。在本文中,我们提出了一个信息论的角度来看,BNN的训练。我们通过信息瓶颈原理分析BNN,并观察到BNN的训练动态与深度神经网络(DNN)的训练动态有很大不同。虽然DNN具有单独的经验风险最小化和表示压缩阶段,但我们的数值实验表明,在BNN中,这两个阶段是同时进行的。由于BNN的表达能力较低,它们倾向于在标签拟合的同时找到有效的隐藏表示。多个数据集的实验支持这些观察结果,我们看到BNN中不同激活函数的行为一致。

英文摘要:

Compact neural networks are essential for affordable and power efficient deep learning solutions. Binary Neural Networks (BNNs) take compactification to the extreme by constraining both weights and activations to two levels, $\{+1, -1\}$. However, training BNNs are not easy due to the discontinuity in activation functions, and the training dynamics of BNNs is not well understood. In this paper, we present an information-theoretic perspective of BNN training. We analyze BNNs through the Information Bottleneck principle and observe that the training dynamics of BNNs is considerably different from that of Deep Neural Networks (DNNs). While DNNs have a separate empirical risk minimization and representation compression phases, our numerical experiments show that in BNNs, both these phases are simultaneous. Since BNNs have a less expressive capacity, they tend to find efficient hidden representations concurrently with label fitting. Experiments in multiple datasets support these observations, and we see a consistent behavior across different activation functions in BNNs.

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。

↑