This paper aims to predict and optimize the behavior of solid particle erosion in polymethyl methacrylate material through artificial neural networks assisted with the implementation of metaheuristic algorithms. For the study, artificial neural networks (ANNs) and the corresponding metaheuristic algorithms, such as genetic algorithm (GA), particle swarm optimization (PSO), whale optimization algorithm (WOA), gray wolf optimizer (GWO), student psychology-based optimization algorithm (SPBO), ant lion optimizer (ALO), teaching learning-based optimization (TLBO), equilibrium optimizer (EO), flow direction algorithm (FDA), and covariance matrix adaptation evolution strategy (CMA-ES) have been implemented along with jellyfish search optimizer (JSO) to predict the erosion rate and erosion velocity of poly methyl methacrylate (PMMA) eroded by white aluminum oxide at different impact angles and particle sizes. It was indicated that ANN and metaheuristic prediction models were suited to the experimental results, affirming that the approach used is a successful one in predicting solid particle erosion of polymeric materials. JSO exhibited the most robust results with the lowest standard deviation of the erosion rate. In the present study, it is observed that the experimental data are in good agreement with what is obtained from the ANN model fed with input parameters optimized using various metaheuristic algorithms. In the experiment, characterization with polynomials was done regarding the aluminum oxide (Al2O3) abrasive particles' erosion behavior on PMMA at three different mesh sizes and three different blasting pressures. These polynomials were used as the fitness functions for the metaheuristic algorithms applied. The results show that the combined use of ANNs and metaheuristics yields even higher prediction accuracy than those provided by each alone. The JSO algorithm has demonstrated excellence in erosion rate and amount and in its ability to optimize over other approaches. The differences between the results of erosion rates from the experimental studies with the erosion rates estimated from ANN and metaheuristic algorithms were discussed in this context. Based on the experimental results, the erosion rate from the tests, in which 60-mesh size Al2O3 particles were used under the pressure level of 3 Bar and impact angle of 45°, is 0.45%. Then, the predicted erosion rate from the JSO algorithm is 0.44%. Thus, this result establishes that the JSO algorithm is a great predictor with high accuracy.