Keynote Talk: Orthogonal Neural Network - An approach for Solving Delay Differential Equations

The neural network approach stands out as a prominent method for solving differential equations, illuminating new avenues of research over the past two decades. We propose an efficient orthogonal neural network with an extreme learning machine algorithm to obtain approximate solution of higher order neutral delay differential equations with variable coefficients. The applicability of this approach ranges from single delay differential equation to a system of delay differential equations. To demonstrate its efficacy, we present numerical simulation for variety of problems highlighting the effectiveness of the proposed methodology.