BriefGPT.xyz
Sep, 2017
神经网络中的过度泛化的去噪自编码器
Denoising Autoencoders for Overgeneralization in Neural Networks
HTML
PDF
Giacomo Spigler
TL;DR
本研究提出了一种使用去噪自编码器计算置信度得分的新方法，并展示了这种置信度得分可以通过确定其局部极大值来正确识别接近训练分布的输入空间的区域，从而解决神经网络模型的过度概括问题。
Abstract
Despite the recent developments that allowed
neural networks
to achieve impressive performance on a variety of applications, these models are intrinsically affected by the problem of
overgeneralization
, due to th
→