Derives explicit excess risk bounds for metric and similarity learning via structured deep ReLU networks that approximate the true metric under hinge loss.
Nearly-tight vc-dimension and pseudodimension bounds for piecewise linear neural networks
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
stat.ML 1years
2024 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Generalization analysis with deep ReLU networks for metric and similarity learning
Derives explicit excess risk bounds for metric and similarity learning via structured deep ReLU networks that approximate the true metric under hinge loss.