Comprehensive Analysis and Research on the Mainstream Algorithm of Machine Learning in Stock Trend Prediction

EasyChair Preprint 9856

Abstract

Stock market is an important part of financial market and closely related to economic development. Various problems of stock price analysis and prediction have always existed with the establishment of financial market. Therefore, this paper uses the historical transaction data of Fenghua Hi-Tech As the research object to forecast and analyze the trend of its rise and fall. Through Support Vector Machine(SVM) model, LGBM model, Random Forest(FM) model to predict the stock trend. Through the empirical study, combined with the prediction chart of each model and the evaluation indexes of stock rise and fall prediction, such as RMSE, MAE, MAPE and R2, the prediction effect and prediction accuracy of the model were demonstrated. Finally, it was concluded that the model based on Random Forest had better prediction accuracy.

Keyphrases: FM model, LGBM model, SVM model, Stock Prediction

@booklet{EasyChair:9856,
  author    = {Xiuyan Zheng and Jiajing Cai and Jiabin Wang and Shangyu Meng and Guangfu Zhang and Jinling Wei},
  title     = {Comprehensive Analysis and Research on the Mainstream Algorithm of Machine Learning in Stock Trend Prediction},
  howpublished = {EasyChair Preprint 9856},
  year      = {EasyChair, 2023}}