Fractional Hermite–Hadamard-Type Inequality for Interval-Valued cr-Convex Functions with Artificial Neural Network Analysis

In this paper, a Hermite–Hadamard-type inequality for interval-valued functions is investigated within the framework of conformable fractional calculus. First, new results are established for cr-convex interval-valued functions by using the conformable fractional integral operator. Through appropriate choices of parameters, the obtained inequality covers and generalizes many known Hermite–Hadamard-type results. To illustrate the theoretical findings, an explicit numerical example is presented together with graphical representations. In addition, an artificial neural network (ANN) model is implemented as a computational approximation tool to investigate the center-radius forms of the left, middle, and right terms of the inequality across a broad parameter domain. The numerical outputs demonstrate the high predictive accuracy of the trained ANN model in approximating the analytical expressions.