A Novel Radial Basis Function Neural Network with High Generalization Performance for Nonlinear Process Modelling

Abstract

A radial basis function neural network (RBFNN), with a strong function approximation ability, was proven to be an effective tool for nonlinear process modeling. However, in many instances, the sample set is limited and the model evaluation error is fixed, which makes it very difficult to construct an optimal network structure to ensure the generalization ability of the established nonlinear process model. To solve this problem, a novel RBFNN with a high generation performance (RBFNN-GP), is proposed in this paper. The proposed RBFNN-GP consists of three contributions. First, a local generalization error bound, introducing the sample mean and variance, is developed to acquire a small error bound to reduce the range of error. Second, the self-organizing structure method, based on a generalization error bound and network sensitivity, is established to obtain a suitable number of neurons to improve the generalization ability. Third, the convergence of this proposed RBFNN-GP is proved theoretically in the case of structure fixation and structure adjustment. Finally, the performance of the proposed RBFNN-GP is compared with some popular algorithms, using two numerical simulations and a practical application. The comparison results verified the effectiveness of RBFNN-GP.

Social network based link correlation using graph neural network with deep learning architectures for feature vectors prediction and classification

“…where p(t i |x i ; W) spatially weighted cross-entropy loss is written as: indicates the probability prediction of a voxel xi after soft-max function in the final output layer as shown in Equation (1).…”

“…By following a similar procedure, we can represent the decomposition of three chain shift operator. Specifically, by utilizing formulas x (2) = S 2 x (1) and expansion (34), we have…”

“…It can be seen as a graph, with each group member represented by a vertex and any associations between them defined by edges. Associations within a group are commonly motivated by internalized common interests 1 . Social networks are particularly dynamic objects because they constantly grow as new vertices and edges are added to the graph.…”

Underwater Aerial Vehicle Networks Based Image Analysis By Deep Learning Architecture Integrated With 5G System

“…In addition to this effect, it's important to note how different underwater picture processing techniques, such sea snowfall, enhance amount of dispersed light [5].To alter the existing colour schemes as well asimprove image for future processing, a pre-processing phase is therefore necessary.It might be approached from two different angles.The goal of image restoration is to recover a wavelet coefficient from an obtained image and a degradation model.The favoured technique, image improvement, makes advantage of contextually relevant individual features to enhance the aesthetics of an image.The main distinction between the two approaches is that, while picture restoration yields more realistic results, it also necessitates the estimation or measurement of a number of factors [6]. Contribution of this research is as follows:…”

Applying radial basis function neural network for comprehending properties of each cluster of fuzzy c-means in coordinates analysis (case study in Iran)

“…In recent decade, with arti cial intelligence's ongoing promotion, data-driven algorithms have been utilized so much due to they consist simple mathematical models. behind them, radial basis function neural network is so popular (Yang et al, 2022). In general, radial basis function neural networks suggest the useful algorithm for non-linear mapping and clustering (Wen et al, 2016).…”