Extensive research has been conducted in the past few years to improve artificial neural network (ANN) through several optimization methods. This paper provides a comparative study of optimization algorithms combined with ANN, such as particle swarm optimization (PSO), artificial bee colony (ABC), backtracking search algorithm (BSA), evolutionary algorithm (EA), and genetic algorithm (GA). This research focuses on optimizing neural networks through the hybridization of optimization algorithms to get the optimal input parameters for the tensile strength of 3D-printed woven glass fiber-reinforced PLA composites. All the optimization techniques showed similar configurations of optimal input parameters with different values of the fitness function. Particle swarm optimization technique (PSO) showed the highest fitness value (39.404), while BSA showed the lowest value of mean squared error (MSE) (0.002518). The hybridization of optimization techniques has been successfully shown to predict the optimum configuration of 3D printing parameters, showcasing the advantages of using these algorithms to enhance the mechanical properties of 3D-printed composites.