BriefGPT.xyz
Apr, 2021
深度先验的神经架构搜索：实现无灾难干扰的连续学习
Neural Architecture Search of Deep Priors: Towards Continual Learning without Catastrophic Interference
HTML
PDF
Martin Mundt, Iuliia Pliushch, Visvanathan Ramesh
TL;DR
使用神经网络结构的神经架构搜索并分析其分类性能，证明深度先验可以使线性分类器效果与全面训练的深度对应结构一致，这种深度先验不需要额外的推理，并在连续学习的扩展上研究了无灾难干扰的递增学习的可能性。
Abstract
In this paper we analyze the
classification performance
of
neural network
structures without parametric inference. Making use of
neural architect
→