Optimization of multimodal trait prediction using Particle Swarm Optimization

Multimodal trait prediction is one of the hardest problems in the domain of Computer Science, Machine Learning, and neural networks. Human traits are subjected to changes in terms of time, situation, place, observer, etc. This paper will try to overcome the problem through the optimization of multimodal trait prediction using Particle Swarm Optimization (PSO) algorithm. Parameter optimization problem based on PSO shown in this paper represents a method that is more efficient for both linear and nonlinear models. The obtained results show that PSO can improve both the prediction of the aggregation model which gives a linear approximation of traits and the nonlinear robust estimation models based on the Huber function.