Networks and the best approximation property
From MaRDI portal
Networks can be considered as approximation schemes. It is well known that multilayer networks of the perceptron type can approximate arbitrarily well continuous functions. A similar result is proven for networks derived from regularization theory and including radial basis functions. From the point of view of approximation theory, however, the property of approximating continuous functions arbitrarily well is not sufficient for characterizing good approximation schemes. More critical is the property of best approximation, that is the presence of an element of minimum distance from the function that has to be approximated. In this paper it is shown that multilayer perceptron networks, of the type used in backpropagation, do not have this property. For regularization networks (in particular radial basis function networks) existence and uniqueness of best approximation immediately derives from the linearity of the theory.
Recommendations
- scientific article; zbMATH DE number 5233923
- scientific article; zbMATH DE number 176253
- Sup-norm approximation bounds for networks through probabilistic methods
- An approximation approach to network information theory
- On optimal random nets
- Approximation algorithms for certain network improvement problems
- Approximability of robust network design
- Approximability of robust network design
- Approximation in Large-Scale Circuit-Switched Networks
- Network approximation of input-output maps and functionals
Cites work
- Applications of the Theory of Boolean Rings to General Topology
- Approximation by superpositions of a sigmoidal function
- Approximation from a curve of functions
- scientific article; zbMATH DE number 3974187 (Why is no real title available?)
- scientific article; zbMATH DE number 4070373 (Why is no real title available?)
- scientific article; zbMATH DE number 4071579 (Why is no real title available?)
- scientific article; zbMATH DE number 4074523 (Why is no real title available?)
- scientific article; zbMATH DE number 3551792 (Why is no real title available?)
- scientific article; zbMATH DE number 3449248 (Why is no real title available?)
- scientific article; zbMATH DE number 3227378 (Why is no real title available?)
- scientific article; zbMATH DE number 3301601 (Why is no real title available?)
- scientific article; zbMATH DE number 3399886 (Why is no real title available?)
- Interpolation of scattered data: distance matrices and conditionally positive definite functions
- Learning representations by back-propagating errors
- The convergence of sequences of rational functions of best approximation
Cited in
(83)- Model predictive control of nonlinear hybrid systems with discrete inputs employing a hybrid fuzzy model
- Classification and identification of mosquito species using artificial neural networks
- Neural networks for control systems - a survey
- From basis functions to basis fields: Vector field approximation from sparse data
- Some extensions of radial basis functions and their applications in artificial intelligence
- On the approximation of functional classes equipped with a uniform measure using ridge functions
- Extending the smoothing spline estimate
- A theory for memory-based learning
- On radial basis function nets and kernel regression: Statistical consistency, convergence rates, and receptive field size
- Using neural networks to aid the diagnosis of breast implant rupture
- Best approximation by linear combinations of characteristic functions of half-spaces.
- Adaptive NN control for a class of strict-feedback discrete-time nonlinear systems
- Approximation of function and its derivatives using radial basis function networks
- Approximation properties of local bases assembled from neural network transfer functions
- Optimised PWL recursive approximation and its application to neuro-fuzzy systems
- Interpolation by ridge polynomials and its application in neural networks
- Approximation theorems for a family of multivariate neural network operators in Orlicz-type spaces
- A neural network method for nonlinear time series analysis
- Provable approximation properties for deep neural networks
- A unified framework for harmonic analysis of functions on directed graphs and changing data
- Function approximation on non-Euclidean spaces
- A Monte Carlo strategy for data-based mathematical modeling
- Modelling and control of nonlinear, operating point dependent systems via associative memory networks
- A neural network controller with PID compensation for trajectory tracking of robotic manipulators
- Using radial basis function networks for function approximation and classification
- Predictive control of air-fuel ratio in aircraft engine on fuel-powered unmanned aerial vehicle using fuzzy-RBF neural network
- Topological properties of the set of functions generated by neural networks of fixed size
- A direct approach for function approximation on data defined manifolds
- Deep learning Gauss-Manin connections
- Best \(k\)-layer neural network approximations
- Stretchy binary classification
- Adaptive neural network control for nonstrict-feedback uncertain nonlinear systems with input delay and asymmetric time-varying state constraints
- Effective implementation to reduce execution time of a low-rank matrix approximation problem
- A survey of randomized algorithms for training neural networks
- Density problem and approximation error in learning theory
- An application of mapping neural networks and a digital signal processor for cochlear neuroprostheses
- On optimal random nets
- Restricted gradient-descent algorithm for value-function approximation in reinforcement learning
- Face and palmprint pixel level fusion and kernel DCV-RBF classifier for small sample biometric recognition
- Machine learning based data retrieval for inverse scattering problems with incomplete data
- Non-parametric identification of homogeneous dynamical systems
- Hierarchical sparse method with applications in vision and speech recognition
- Quick approximation of bivariate functions
- Model selection approaches for non-linear system identification: a review
- Regularization algorithms for learning that are equivalent to multilayer networks
- Nonlinear one-step-ahead predictive mean control of bounded dynamic stochastic systems with guaranteed stability
- Consistency of learning algorithms using Attouch-Wets convergence
- Robot competence development by constructive learning
- Multiobjective Metamodel–Assisted Memetic Algorithms
- Neuro-fuzzy control of complex manufacturing processes
- A multilayer nodal link perceptron network with least squares training algorithm
- Feedforward control of a class of hybrid systems using an inverse model
- scientific article; zbMATH DE number 637314 (Why is no real title available?)
- On function recovery by neural networks based on orthogonal expansions
- A constraint sampling approach for multi-stage robust optimization
- RBF interpolation of boundary values in the BEM for heat transfer problems
- Bayesian ARTMAP for regression
- Experimental design and model construction algorithms for radial basis function networks
- UNIFIED FRAMEWORK FOR MLPs AND RBFNs: INTRODUCING CONIC SECTION FUNCTION NETWORKS
- Sup-norm approximation bounds for networks through probabilistic methods
- One-to-one mapping and its application to neural networks based control systems design
- Universal neural controllers and stability analysis
- Event-triggered adaptive decentralised control for switched interconnected nonlinear systems with unmodeled dynamics and full state constraints
- Sigmoidal FFANN’s and the best approximation property
- Approximation of 3D trapezoidal fuzzy data using radial basis functions
- Event-triggered cooperative control for high-order nonlinear multi-agent systems with finite-time consensus
- Observer‐based adaptive neural networks control for Markovian jump nonlinear systems with partial mode information and input saturation
- Nonclosedness of sets of neural networks in Sobolev spaces
- Event-triggered adaptive neural networks control for fractional-order nonstrict-feedback nonlinear systems with unmodeled dynamics and input saturation
- Reinforcement learning‐based optimized backstepping control for strict‐feedback nonlinear systems subject to external disturbances
- Hierarchical regularization networks for sparsification based learning on noisy datasets
- A Bayesian regularization network approach to thermal distortion control in 3D printing
- Measuring Complexity of Learning Schemes Using Hessian-Schatten Total Variation
- Native spaces and generalization of Wu functions
- On the existence of optimal shallow feedforward networks with ReLU activation
- On the existence of global minima and convergence analyses for gradient descent methods in the training of deep neural networks
- Random projection neural networks of best approximation: convergence theory and practical applications
- Singular parameters and missing limits in neural PDE solvers
- Comparing fixed and variable-width Gaussian networks
- Design and performance analysis of deterministic learning of sampled-data nonlinear systems
- Establishing criteria to ensure successful feedforward artificial neural network modelling of mechanical systems
- Solving a system of nonlinear integral equations by an RBF network
- A new kernel-based approach for linear system identification
This page was built for publication: Networks and the best approximation property
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q751636)
