A Literature Review on Machine Learning Techniques and Strategies Applied to Stock Market Price Prediction

Patil, Pankaj Rambhau; Parasar, Deepa; Charhate, Shrikant

doi:10.1007/978-981-16-3915-9_10

Cited by 5 publications

(2 citation statements)

References 66 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Numerous fields, including stock price prediction, have benefited from the use of machine learning and deep learning [7]. Mehar Vijh (2019) predicted the closing prices of five firms in various operating sectors the next day using Artificial Neural Network and Random Forest methods [8]. RMSE and MAPE, two common strategic measures, are used to assess these models.…”

Section: Related Researchmentioning

confidence: 99%

“…Taking the autocorrelation relationship into consideration, I construct a model (5). And the estimated model is E(ln (AAPL t )) = −0.5744 + 0.0982 * ln(NASDAQ t ) + 0.9259 * ln(AAPL t−1 ) (8) It is surprising that the R square for the model is up to 0.9972, which means that a 99.72% variation of AAPL stock price can be explained by this log-log model with autocorrelation term. It performs very well.…”

Section: Model Developmentmentioning

confidence: 99%

See 1 more Smart Citation

Using the Nasdaq Index to Predict AAPL Price by Linear Regression Analysis

Jin¹,

Ma²

2023

BCPBM

View full text Add to dashboard Cite

In this project, we want to predict AAPL’s stock price by the NASDAQ index by the regression model. The dependent variable is AAPL’s stock price, and the independent variable is the NASDAQ index. First, we do some descriptive statistics for the two variables and measure the distribution from the central tendency, variation tendency, and distribution to acknowledge the character of distributions. Based on the strong linear relationship between AAPL stock price and the NASDAQ index, we constructed a simple linear regression model. Considering the scale of the two variables, we tried the other three models with log transformation. And then, it shows that the log-log model has the best performance. However, in the residual analysis of the log-log model, it shows an autocorrelation in the residual, then we generate a new variable that is the one-order term for AAPL and add it into the model, and it surprisingly performs very well, whose R square is up to 99.72%. Therefore, we think combining the linear relationship with the market and the autocorrelation itself can construct a good model, and it can apply to predict much other stock's prices in the market.

show abstract

Section: Related Researchmentioning

confidence: 99%

Section: Model Developmentmentioning

confidence: 99%