Particle Swarm Optimization based support vector machine for damage level prediction of non-reshaped berm breakwater

- Particle Swarm Optimization (PSO) is used to optimize the support vector machine (SVM).
- Models are trained on the data set obtained from experimental data.
- PSO-SVM with polynomial kernel function performs better than other kernel functions.
- Different soft computing models results are compared.
- PSO-SVM is computationally efficient as compared to ANFIS.