ExSpliNet: An interpretable and expressive spline-based neural network
From MaRDI portal
Abstract: In this paper we present ExSpliNet, an interpretable and expressive neural network model. The model combines ideas of Kolmogorov neural networks, ensembles of probabilistic trees, and multivariate B-spline representations. We give a probabilistic interpretation of the model and show its universal approximation properties. We also discuss how it can be efficiently encoded by exploiting B-spline properties. Finally, we test the effectiveness of the proposed model on synthetic approximation problems and classical machine learning benchmark datasets.
Cites work
- A practical guide to splines.
- An identity for spline functions with applications to variation diminishing spline approximation
- Approximation by series of sigmoidal functions with applications to neural networks
- Approximation by superpositions of a sigmoidal function
- Approximation in FEM, DG and IGA: a theoretical comparison
- Breaking the curse of dimensionality with convex neural networks
- DeepXDE: a deep learning library for solving differential equations
- Error bounds for deep ReLU networks using the Kolmogorov-Arnold superposition theorem
- Explicit error estimates for spline approximation of arbitrary smoothness in isogeometric analysis
- Foundations of spline theory: B-splines, spline approximation, and hierarchical refinement
- scientific article; zbMATH DE number 6378127 (Why is no real title available?)
- scientific article; zbMATH DE number 3147701 (Why is no real title available?)
- scientific article; zbMATH DE number 3860199 (Why is no real title available?)
- scientific article; zbMATH DE number 5190601 (Why is no real title available?)
- scientific article; zbMATH DE number 47282 (Why is no real title available?)
- scientific article; zbMATH DE number 1348793 (Why is no real title available?)
- scientific article; zbMATH DE number 1928659 (Why is no real title available?)
- Multilayer feedforward networks are universal approximators
- Neural random forests
- New error bounds for deep ReLU networks using sparse grids
- Nonlinear system identification with regularized tensor network B-splines
- On a constructive proof of Kolmogorov's superposition theorem
- Physics-informed neural networks: a deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations
- Sharp error estimates for spline approximation: explicit constants, n-widths, and eigenfunction convergence
Cited in
(9)- Interpretable Generative Adversarial Networks With Exponential Function
- Spline representation and redundancies of one-dimensional ReLU neural network models
- A comprehensive and FAIR comparison between MLP and KAN representations for differential equations and operator networks
- Optimal linear B-spline approximation via Kolmogorov superposition theorem and its applications
- On the expressivity of the ExSpliNet KAN model
- DeepOKAN: deep operator network based on Kolmogorov Arnold networks for mechanics problems
- Neural tangent kernel analysis to probe convergence in physics-informed neural solvers: PIKANs vs. PINNs
- A Kolmogorov-Arnold neural model for cascading extremes
- Physics informed Kolmogorov-Arnold neural networks for dynamical analysis via efficient-KAN and WAV-KAN
This page was built for publication: ExSpliNet: An interpretable and expressive spline-based neural network
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6398139)
