Time Series Analysis and Predictions of Stock Market Data Using Deep Learning Techniques

EasyChair Preprint 12464

Abstract

The stock market is a complex and dynamic system that is affected by many factors, from economic indicators to political events to the opinions of businessmen. Traditional time analysis methods have been used for decades, but recent advances in deep learning have yielded great results in modeling and predicting complex time products. In this study, we use short-term temporal (LSTM) neural network and convolutional neural network (CNN) to analyze and predict business data. We compare the performance of this model with traditional time-based models such as ARIMA and exponential smoothing. Our results show that deep learning models can outperform traditional models in stock market forecasting, with LSTM models achieving the best performance. We also show that integration with other sources, such as news and social insights, can improve the accuracy of predictions. Our findings suggest that deep learning techniques can be valuable tools for investors and analysts who want to make informed decisions based on stock market data. However, we caution that the complexity and inconsistencies of the stock market make accurate predictions difficult, and we recommend using this formula as part of a good investment

Keyphrases: CNN, LSTM, decision trees, deep learning

@booklet{EasyChair:12464,
  author    = {Chintalacheruvu V N Bharadwaj and Dhaval Nimavat},
  title     = {Time Series Analysis and Predictions of Stock Market Data Using Deep Learning Techniques},
  howpublished = {EasyChair Preprint 12464},
  year      = {EasyChair, 2024}}