Prediction of share market stock price using long short-term memory and compare accuracy with logistic regression algorithm

Improving the precision and pinpointing of stock pricing predictions employing Long Short-Term Memory (LSTM) compared to Logistic Regression (LR) is the primary objective of this study paper. Research Tools and Procedures: This paper’s dataset demonstrates the approach’s efficacy using the publicly accessible dataset from the National Stock Exchange (NSE). The sample size for improving the accuracy and precision of share market stock price predictions was 280 (140 in Group 1 and 140 in Group 2). The computation was done using G-power 0.8 with alpha and beta values of 0.05 and 0.2, respectively, and a confidence interval of 95%. Both LSTM and LR use a large number of samples (N=10) to forecast stock prices on the stock market with a high degree of accuracy and precision. In comparison to the LR’s 82.51% accuracy rate, the LSTM classifier achieves a far better rate of 93.94%. A p-value of 0.971 indicates that the research is not statistically significant (p>0.05). For the purpose of improving accuracy and precision in predicting the marketplace for stock prices, LSTM produces superior results in terms of prediction accuracy than LR.