Kantorovich Tipi Maks-Min Yapay Sinir Ağı Operatörlerinin Lp-Yakınsaması

Özet (abstract)

In this work, we study the Kantorovich variant of max-min neural network operators, in which the operator kernel is defined in terms of sigmoidal functions. Our main aim is to demonstrate the Lp-convergence of these nonlinear operators for 1≤p<∞, which makes it possible to obtain approximation results for functions that are not necessarily continuous. In addition, we will derive quantitative estimates for the rate of approximation in the Lp-norm. We will provide some explicit examples, studying the approximation of discontinuous functions with the max-min operator, and varying additionally the underlying sigmoidal function of the kernel. Furthermore, we numerically compare the Lp-approximation error and computational time complexity of our operator with the Kantorovich variants of other prominent neural network operators. As an application, we illustrate that the Kantorovich max-min operator can offer advantages over the classical variant and the Kantorovich max-product operators when filtering noisy functions, such as biomedical ECG signals.

Yazarların özeti; kaynağından alınmıştır. Numerical Functional Analysis and Optimization, 2026 · DOI ↗