ChebNet builds stable deep RePU networks via hierarchical Chebyshev approximations in the frequency domain, matching power-series approximation quality while improving stability and enabling fine-tuning.
PowerNet: Efficient representations of polynomials and smooth functions by deep neural networks with rectified power units
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2019 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
ChebNet: Efficient and Stable Constructions of Deep Neural Networks with Rectified Power Units via Chebyshev Approximations
ChebNet builds stable deep RePU networks via hierarchical Chebyshev approximations in the frequency domain, matching power-series approximation quality while improving stability and enabling fine-tuning.