Neural networks always beat the constant predictor in population-level regression when a relationship exists, whereas any identifiable smooth parametric model can be reduced to predicting the mean for some distribution.
Universal approximation bounds for superpositions of a sigmoidal function
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
math.ST 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Non-identifiability distinguishes Neural Networks among Parametric Models
Neural networks always beat the constant predictor in population-level regression when a relationship exists, whereas any identifiable smooth parametric model can be reduced to predicting the mean for some distribution.