调参如何影响神经网络复杂度与抗干扰能力
Assessing Simplification Levels in Neural Networks: The Impact of Hyperparameter Configurations on Complexity and Sensitivity
- 通过调整激活函数、层数和学习率，研究模型输出复杂度变化
- 发现不同配置下网络输出复杂度差异显著，敏感性随之改变
- 适合关注模型可解释性与鲁棒性的研究人员参考
本文通过实验研究不同超参数配置下神经网络的简化特性，重点考察了Lempel-Ziv复杂度与敏感性的影响。通过调整激活函数、隐藏层数量和学习率等关键超参数，分析其对网络输出复杂度及对输入扰动的鲁棒性的影响。实验基于MNIST数据集进行，旨在揭示超参数、复杂度与敏感性之间的关系，为理解神经网络中的这些概念提供更深入的理论基础。
原文摘要 · Abstract (English)
This paper presents an experimental study focused on understanding the simplification properties of neural networks under different hyperparameter configurations, specifically investigating the effects on Lempel Ziv complexity and sensitivity. By adjusting key hyperparameters such as activation functions, hidden layers, and learning rate, this study evaluates how these parameters impact the complexity of network outputs and their robustness to input perturbations. The experiments conducted using the MNIST dataset aim to provide insights into the relationships between hyperparameters, complexity, and sensitivity, contributing to a deeper theoretical understanding of these concepts in neural networks.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。