作者: Shenghuo Zhu , Yuanqing Lin , Kai Yu , Zhang Tong
DOI:
关键词: Functional approximation 、 Linear network coding 、 Nonlinear system 、 Improved performance 、 Theoretical computer science 、 Neural coding 、 Coding (social sciences) 、 Mathematics
摘要: This paper proposes a principled extension of the traditional single-layer flat sparse coding scheme, where two-layer scheme is derived based on theoretical analysis nonlinear functional approximation that extends recent results for local coordinate coding. The approach can be easily generalized to deeper structures in hierarchical multiple-layer manner. Empirically, it shown deep yields improved performance benchmark datasets.