Multilayer feedforward networks are universal approximators
Neural Networks
Approximation capabilities of multilayer feedforward networks
Neural Networks
A resource-allocating network for function interpolation
Neural Computation
Neural Networks: Tricks of the Trade, this book is an outgrowth of a 1996 NIPS workshop
Letters: Convex incremental extreme learning machine
Neurocomputing
Efficient agnostic learning of neural networks with bounded fan-in
IEEE Transactions on Information Theory - Part 2
Universal approximation bounds for superpositions of a sigmoidal function
IEEE Transactions on Information Theory
Objective functions for training new hidden units in constructive neural networks
IEEE Transactions on Neural Networks
Universal approximation using incremental constructive feedforward networks with random hidden nodes
IEEE Transactions on Neural Networks
Hybrid extreme rotation forest
Neural Networks
Applications of Hybrid Extreme Rotation Forests for image segmentation
International Journal of Hybrid Intelligent Systems
Hi-index | 0.01 |
In this paper, we systemically investigate several convex incremental feedforward neural networks. Firstly, we prove the universal approximation and the convergence rate of a generalized convex incremental (GCI) structure, which provides us a wider parameter selection. Second, according to the convergence rate proof of GCI, we further prove the convergence rate of a best convex incremental (BCI) structure, moreover its proof also illustrates that BCI can achieve a better generalization performance than GCI. But we should note that the hidden neurons of BCI and GCI both are constructed on the maximum principle (not random). Next, we introduce the random neuron conception based on CI-ELM (convex incremental extreme learning machines), and further propose an alternative algorithm (improved CI-ELM, ICI-ELM) between CI-ELM and BCI, which removes the ''useless'' neurons in CI-ELM and improves the efficiency of neural networks. ICI-ELM randomly generates a group of parameters, among which we determine the best parameters leading to the smallest residual error. Therefore ICI-ELM can achieve a faster convergence rate than CI-ELM, meanwhile it still retains the same convergence rate as BCI. On the other hand, ICI-ELM also provides an alternative scheme to replace conventional gradient methods, which are only suitable for differential functions and often achieves local minima. The experimental results based on several benchmark regression problems support our claims.