RVM based on PSO for Groundwater Level Forecasting

Relevance Vector Machine (RVM) is a novel kernel method based on Sparse Bayesian, which has many advantages such as its kernel functions without the restriction of Mercer’s conditions, the relevance vectors automatically determined. In this paper, a new RVM model optimized by Particle Swarm Optimization (PSO) is proposed, and it is applied to groundwater level forecasting. The simulation experiments demonstrate that the proposed method can reduce significantly both relative mean error and root mean squared error of predicted groundwater level. Moreover, the model achieved is much sparser than its counterpart, so the RVM based on PSO is applicable and performs well for groundwater data analysis.