Response to letter to the editor entitled: “Correlation analysis of the triglyceride glucose index and heart failure with preserved ejection fraction in essential hypertensive patients”

A study on vegetable stocking-pricing decision-making in supermarkets based on regression analysis and planning models

“…RBFNN is a three-layer feed-forward neural network that can globally approximate any nonlinear relationship with arbitrary accuracy. It has the advantages of fast learning convergence, no local minima problem, and strong generalisation ability, and is widely used in the fields of nonlinear function approximation, pattern recognition, system control, and prediction.The RBFNN structure consists of an input layer, a hidden layer, and an output layer [4].The RBFNN input layer transmits the signals, while the radial basis function maps the input layer to the hidden layer, and a simple linear weighting is used between the hidden layer and the output layer. The performance of the RBFNN is a function of its extended Spread; the larger the Spread, the smoother the function fit is, but the larger the approximation error, thus requiring more hidden layer neurons.The smaller the Spread, the more accurate the function approximation is, and the smoothing of the approximation process decreases, which reduces the performance of the network [5].…”