Intelligent heuristic computing paradigm: A novel strategy to singular differential difference equations analysis

Intelligent heuristic computing paradigm: A novel strategy to singular differential difference equations analysis

n the presented study, a novel intelligent heuristic computing paradigm based on artificial intelligence is introduced for solving linear singular differential difference equations (SDDEs). Unsuper-vised artificial neural networks (ANNs) with universal function approximation capabilities are employed to establish a mathematical model for the problem, incorporating a mean squared error function. The design parameter training for ANN models involves utilizing the global search capabilities of a genetic algorithm (GA), an effective local search through an interior point approach (IPA), and hybridization of GA-IPA. The precision, reliability, and robustness of the proposed methodology are endorsed through comprehensive comparative analysis against exact solutions and previously reported findings in the lit-erature. Statistical assessments of the outcomes are employed to confirm the accuracy and convergence of the design approach. Multiple independent runs of these algorithms are also conducted and compared against approximate numerical solutions to ensure accuracy and convergence.