International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 10 Issue: 04 | Apr 2023
p-ISSN: 2395-0072
www.irjet.net
Predicting Stock Market Prices with Sentiment Analysis and Ensemble Learning Techniques: A Hybrid Approach Alister Rodrigues1, Sumedh Salve2, Tanfaiz Shaikh3, Priyanka Bhilare4 1,2,3Student, Computer Engineering, MCT’s Rajiv Gandhi Institute of Technology, Mumbai University 4Professor, Dept. of Computer Engineering, MCT’s Rajiv Gandhi Institute of Technology, Mumbai University
---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - Stock market price prediction has been a
analysis with machine learning models to predict stock prices. This paper proposes a hybrid approach that incorporates sentiment analysis with ensemble learning techniques to predict stock market prices. The proposed approach seeks to capture the impact of external factors such as news and social media on the stock market and improve the prediction accuracy of the model. The ensemble learning technique serves to reduce the effect of overfitting and increase the robustness of the model. The balance of the paper is organized as follows: Section 2 provides a literature review of the related work on stock market prediction using sentiment analysis and machine learning models. Section 3 describes the proposed hybrid approach in detail, including the sentiment analysis component and ensemble learning techniques. Section 4 presents the results of the proposed approach and Section 5 concludes the paper.
challenging task for financial analysts and investors. With the rapid development of social media and news platforms, sentiment analysis has gained popularity as a tool for predicting stock prices. This paper proposes a hybrid approach that incorporates sentiment analysis with ensemble learning techniques to predict stock market prices. The proposed approach consists of four main steps: (1) sentiment analysis of news and social media data related to a particular stock; (2) fetching historical stock data for the said stock; (3) feature extraction using various technical indicators; and (4) ensemble learning using a combination of multiple machine learning models. The proposed approach was evaluated based on the stock prices of five key companies in the technology industry. The results showed that the hybrid approach outperformed individual machine learning models and traditional time-series forecasting methods in terms of accuracy and consistency. The ensemble learning technique aided to reduce the effect of overfitting and increase the robustness of the model. The sentiment analysis component contributed to enhancing the prediction accuracy by providing insights into the market's sentiment towards a particular stock. Ultimately, the research project demonstrates the potential of using sentiment analysis and ensemble learning techniques to predict stock market prices. The proposed approach can be used by financial analysts and investors to make informed decisions and mitigate the risks associated with stock investments.
2. LITERATURE REVIEW A. Literature Review 1)“Bankruptcy Prediction for Credit Risk Using Neural Networks: A Survey and New Results” by Amir F. Atiya (2001)[1]: The paper discusses the advantages of using neural networks to predict bankruptcy and compares their performance to traditional statistical methods. It also examines the various input and output variables used in bankruptcy prediction models, including financial ratios, market data, and macroeconomic variables. Finally, it presents new results from a study that employs a neural network to predict bankruptcy using financial ratios as input variables. The paper highlights the advantages of using neural networks in bankruptcy prediction and provides insights into the various inputs and output variables used in bankruptcy prediction models.
Keywords: Stock market, price prediction, social media, ensemble learning, market sentiment, hybrid approach
1.INTRODUCTION With the rise of social media and news platforms, there has been a growing interest in using sentiment analysis to predict stock prices. Sentiment analysis is a technique used to extract subjective information from text data, such as news articles and social media posts, to ascertain the sentiment or opinion of the author towards a particular topic. In recent years, machine learning models have been extensively used to predict stock prices using historical stock market data and technical indicators. However, these models do not take into account the impact of external factors such as news and social media on the stock market. To address this limitation, researchers have proposed hybrid approaches that integrate sentiment
© 2023, IRJET
|
Impact Factor value: 8.226
2)“Stock Market Prediction Using LSTM Recurrent Neural Network” by Adil MOGHAR ,Mhamed HAMICHE (2020) [2] : Machine learning techniques, particularly recurrent neural networks (RNNs), have been popular for stock market prediction. The Long Short-Term Memory (LSTM) architecture of RNNs has been particularly popular due to its ability to model long-term dependencies and manage variable-length sequences of data. Several studies have investigated the use of LSTM RNNs for stock market prediction, and the results have been promising. One study
|
ISO 9001:2008 Certified Journal
|
Page 390