Linearly augmented neural network

James G. Droppo,Pegah Ghahremani, Michael L. Seltzer

user-5f163cde4c775ed682f575fe(2020)

引用 1|浏览18
暂无评分
摘要
A neural network is structured to connect the input values of an input set, at each level, to that level's output using a linear bypass connection. The linear bypass connection passes the input values, to the output, without applying a non-linear function to them.
更多
查看译文
关键词
Connection (mathematics),Function (mathematics),Artificial neural network,Set (abstract data type),Control theory,Mathematics
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要