Performance Analysis of Predicting LIC Stock Price using Lasso Regression Compared with Random Forest Regression

With regard to the objective of the entire study, the goal is to improve the accuracy of predicting stock price using Lasso Regression in comparison to Random Forest Regression. The Novel K-Nearest Neighbors and Support Vector Machine Algorithm are the two groups that are the emphasis of this research activity. Each of these groups is carried out ten times. For the purpose of calculating samples, the Clincal software was utilized. The alpha value was determined to be 0.05, and the confidence interval was set at 95%. We took into consideration the stock price prediction dataset that was available on kaggle.com and had a size of 12617. With a value of p=0.000 (Independent sample t-test p<0.05), the SPSS statistical analysis prediction of stock price demonstrates that there is a statistically significant difference between the Lasso regression algorithm and Random Forest Regression. This data indicates that the Lasso regression algorithm is superior to Random Forest Regression. For the Random Forest regression technique it obtained a percentage accuracy of 84%, while in the new improved Lasso regression approach it records a greater level of accuracy of 93%. The comparison of Random Forest Regressions and Lasso Regressions in the progress of predicting the stock price analysis shows that the latter was substantially more effective. This was revealed in the accomplishment of the current research development project.