Optimization on ELM network using Particle swarm Optimization Algorithms and OSELM to predict the industry index in Tehran Stock Exchange

There have always been two approaches to forecasting in financial markets: traditional and intelligent approaches. In the traditional method, this forecasting is based on statistical models and in the intelligent method is based on artificial intelligence models. Traditional methods mainly use linear patterns to model market behavior, while the main advantage of smart models is the ability to learn and model nonlinear behaviors in the market. It has always been a question of which methods can better model market behavior, and despite the many models that have been proposed for forecasting, there is still an attempt to build a model that can use more effective variables for forecasting. Continues to be able to take into account factors such as time, risk and return. In this research, we have used the neural network to predict the industry index. This is done by ELM neural network using two optimization methods OSELM and PSO. The results show that the prediction accuracy of these two methods is not significantly different from each other, but in terms of execution time, the OSELM neural network algorithm has performed much better and faster.