Prediction and Data Analysis of Price Based on PSO with Extreme Learning Machine Algorithm and Particle Swarm Optimization
Huanyun Chen, Weiming Zhao · 2022
The changes in the trading market are affected by many factors, and traditional forecasting methods are more and more difficult to meet people's needs. In order to improve the accuracy of prediction, this paper proposes an extreme learning machine algorithm based on particle swarm optimization for prediction and analysis. First, the fluctuation data of previous years are collected, and the closing price of the stock is determined as an experiment and simulated to establish a learning sample, and then the learning sample is input. Based on the online extreme learning machine model optimized by particle swarm optimization, the prediction model is established. The results show that the optimized extreme learning machine algorithm, compared with models such as ELM, PSO-BP and DE-ELM, has better surface prediction accuracy and fitting effect, and can more accurately describe the change trend, which is more efficient than other methods.