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 2103.09716cs.CVcs.LG
收藏

基于拓扑熵计算的CNN装置性能定量评价

Quantitative Performance Assessment of CNN Units via Topological Entropy Calculation

Tsinghua University(清华大学)

Yang Zhao, Hao Zhang

摘要:

识别各个网络单元的状态对于理解卷积神经网络(CNN)的机制至关重要。然而,可靠地给出单元状态的一般指示仍然具有挑战性,特别是对于不同网络模型中的单元。为此,我们提出了一种新的方法,用于定量澄清单个单元的状态,在CNN使用代数拓扑工具。单元状态通过定义的基于拓扑的熵(称为特征熵)的计算来指示,特征熵测量隐藏在类别单元中的全局空间模式的混乱程度。通过这种方式,特征熵可以为不同网络中的单元提供准确的状态指示,具有不同的情况,如权重重新缩放操作。此外,我们还发现,特征熵随着层的加深而降低,并且在训练过程中几乎与损失同时发生。我们表明,通过调查的特征熵的单元只有训练数据,它可以给不同的泛化能力的网络之间的区别,从功能表示的有效性的观点。

英文摘要:

Identifying the status of individual network units is critical for understanding the mechanism of convolutional neural networks (CNNs). However, it is still challenging to reliably give a general indication of unit status, especially for units in different network models. To this end, we propose a novel method for quantitatively clarifying the status of single unit in CNN using algebraic topological tools. Unit status is indicated via the calculation of a defined topological-based entropy, called feature entropy, which measures the degree of chaos of the global spatial pattern hidden in the unit for a category. In this way, feature entropy could provide an accurate indication of status for units in different networks with diverse situations like weight-rescaling operation. Further, we show that feature entropy decreases as the layer goes deeper and shares almost simultaneous trend with loss during training. We show that by investigating the feature entropy of units on only training data, it could give discrimination between networks with different generalization ability from the view of the effectiveness of feature representations.

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。

↑