A Hybrid Model for Stock Price Prediction using Machine Learning Techniques with CNN

Predicting the stock market can be a great tool for both long-term and short-term investors to plan and book profits, or to stop losses earlier. Forecasting accuracy is the most crucial factor to consider when choosing a forecasting method. In order to forecast stock markets, we used one of the most...

Full description

Saved in:
Bibliographic Details
Published in2021 5th International Conference on Information Systems and Computer Networks (ISCON) pp. 1 - 6
Main Authors Sarvesh, S., Sidharth, R. V., Vaishnav, V., Thangakumar, J, Sathyalakshmi, S
Format Conference Proceeding
LanguageEnglish
Published IEEE 22.10.2021
Subjects
algo-trading
Convolutional neural networks
LSTM
Machine learning
Memory management
Predictive models
price prediction
Recurrent neural networks
stock market
Time series analysis
Online AccessGet full text

Cover

More Information
Summary:Predicting the stock market can be a great tool for both long-term and short-term investors to plan and book profits, or to stop losses earlier. Forecasting accuracy is the most crucial factor to consider when choosing a forecasting method. In order to forecast stock markets, we used one of the most common recurrent neural networks: LSTM, along with it, Convolutional Neural Network (CNN) is also used. Since the prediction of stocks cannot be easily specified, it can be separated into two parts: simple analysis (sales, revenue, income, etc.) and technical analysis (historical price, VWAP, etc.). This means multiple variables can affect stock price trends, but here we have drawn a predictive time series on the historic price of a given stock. LSTM can quickly process a whole data series and adds a memory cell, which allows the network to link memories and feedback remotely efficiently. In this example we have generated a series of sequences in order to use time steps to predict a given price.
DOI:10.1109/ISCON52037.2021.9702382