Learning robust features by extended generative stochastic networks

作     者：Da Teng Xiao Song Guanghong Gong Junhua Zhou 

作者机构：School of Automation Science and Electrical Engineering Beihang UniversityP.R.China State Key Laboratory of Intelligent Manufacturing System TechnologyP.R.China 

出 版 物：《International Journal of Modeling, Simulation, and Scientific Computing》 (建模、仿真和科学计算国际期刊（英文）)

年 卷 期：2018年第9卷第1期

页      面：135-145页

核心收录：

学科分类：0711[理学-系统科学] 07[理学] 08[工学] 081101[工学-控制理论与控制工程] 0811[工学-控制科学与工程] 071102[理学-系统分析与集成] 081103[工学-系统工程] 

主　　题：Machine learning deep learning neural networks adversarial examples 

摘      要：Deep neural networks have achieved state-of-the-art performance on many object recognition tasks,but they are vulnerable to small adversarial *** this paper,several extensions of generative stochastic networks(GSNs)are proposed to improve the robustness of neural networks to random noise and adversarial *** results show that compared to normal GSN method,the extensions using adversarial examples,lateral connections and feedforward networks can improve the performance of GSNs by making the models more resistant to overfitting and noise.