A novel approach to improve accuracy in stock price prediction using gradient boosting machines algorithm compared with Random Forest algorithm

The work’s goal is to use Gradient Boosting Machines Algorithm to more accurately anticipate stock prices regarding the stock values when compared to the Random Forest Algorithm. Materials and Methods For the purpose of accurately forecasting stock values, iterations of the Gradient Boosting Machines (GBM) and Random Forest algorithms with sample sizes of 20 and 20 were performed, respectively. The Gradient Boosting Machines Algorithm’s Novel Loss Function, which is based on prior stock values, helps to reduce the total prediction error. Results: When compared to the accuracy of the Random Forest Method (83.6%), the GBM algorithm has a much higher accuracy (92.3%). With a significance score of 0.000 (p