AN APPLICATION OF NONLINEAR LEAST SQUARES SUPPORT VECTOR MACHINE USING PARTICLE SWARM OPTIMIZATION TECHNIQUE

摘要

In this paper, we have studied the versatile Nonlinear Least Squares Support Vector Machine (LS-S VM) model. The PBticle Swarm Optimization (PSO), which is a very efficient population-based global stochastic optimization technique is (oployed to estimate the hyper-parameters of this model. Subsequently, as an illustration, the methodology is applied to forecast all-India monthly rainfall time-series data. SAS and MATLAB software packages are used for carrying out the data malysis. Superiority of thisapproach over the Seasonal Autoregressive Integrated Moving Average (SARIMA) model is demonstrated for the data under consideration.