Relation between weight size and degree of over-fitting in neural network regression

DOI10.1016/j.neunet.2007.11.001zbMath1254.62081OpenAlexW2074472353WikidataQ47251599 ScholiaQ47251599MaRDI QIDQ1931976

Kenji Fukumizu, Katsuyuki Hagiwara

Publication date: 17 January 2013

Published in: Neural Networks (Search for Journal in Brave)

Full work available at URL: https://doi.org/10.1016/j.neunet.2007.11.001

zbMATH Keywords

likelihood ratio statisticsdegree of over-fittingneural network regressionweight size

Mathematics Subject Classification ID

Ridge regression; shrinkage estimators (Lasso) (62J07) General nonlinear regression (62J02) Neural networks for/in biological studies, artificial life and related topics (92B20)

Related Items

Asymptotics for regression models under loss of identifiability

Cites Work

• Unnamed Item
• Unnamed Item
• Unnamed Item
• Extremes and related properties of random sequences and processes
• Decision theoretic generalizations of the PAC model for neural net and other learning applications
• Likelihood ratio of unidentifiable models and multilayer neural networks
• On the Problem in Model Selection of Neural Network Regression in Overrealizable Scenario
• The sample complexity of pattern classification with neural networks: the size of the weights is more important than the size of the network
• On the Asymptotic Distribution of the Least-Squares Estimators in Unidentifiable Models
• Neural Network Learning
• An iterated logarithm law for the maximum in a stationary gaussian sequence
• Convergence of stochastic processes