Content area

Abstract

Significant strides have been made in researching stock market trend forecasting and enhancing predictive performance through algorithmic trading methodologies. The effectiveness of automatic trading algorithms has been significantly increased by the incorporation of Artificial Intelligence (AI) into prediction frameworks. However, prevailing research predominantly focuses on daily stock values, projecting trends for subsequent days or weeks. It is widely acknowledged that accurately predicting the precise magnitude of stock price fluctuations is challenging; instead, emphasis is placed on forecasting the direction of price movements to attain superior predictive accuracy. A prevalent limitation of utilizing daily stock values for forecasting daily or weekly price movements is the susceptibility to global market news, which can significantly influence subsequent market behaviors. Addressing this concern necessitates the development of strategies aimed at mitigating the impact of daily market fluctuations, thereby enhancing the utility of predictions. An evident research gap lies in exploring the feasibility of employing reduced time intervals, such as 5-minute intervals, for intraday trading. This study endeavors to fill this void by evaluating the integration of AI in intraday trading using 5-minute stock price intervals. The research focuses on selected stocks trading on the National Stock Exchange (NSE) of India, chosen for their higher beta values and volumes. Leveraging Long-Short Term Memory Networks (LSTM) and Bidirectional Long-Short Term Memory (BiLSTM) models, machine learning techniques are employed to predict price movements within 5-minute intervals. Historical data spanning over two years serves as the training set, with the accuracy of future predictions evaluated against designated test data. The directional accuracy of the predictions exceeded 55%, reaching up to 60% for certain stocks. Furthermore, the study devised a strategic amalgamation of ML predictions with algorithmic logic, resulting in significantly improved profitability ranging from 75% to over 80%. These findings underscore the efficacy of combining ML predictions with established algorithmic strategies, demonstrating substantially higher profitability compared to their application.

Details

1010268
Business indexing term
Title
Hybrid Strategies: Enhancing Intra-Day Trading in India via Machine Learning, Signal Processing, and Market Indicators
Number of pages
169
Publication year
2025
Degree date
2025
School code
4958
Source
DAI-A 87/3(E), Dissertation Abstracts International
ISBN
9798293813285
Committee member
Srivastava, Shishir; Dutta, Kirti
University/institution
American Business Management and Technology College (Switzerland)
Department
Managment
University location
Switzerland
Degree
D.B.A.
Source type
Dissertation or Thesis
Language
English
Document type
Dissertation/Thesis
Dissertation/thesis number
32241332
ProQuest document ID
3247677686
Document URL
https://www.proquest.com/dissertations-theses/hybrid-strategies-enhancing-intra-day-trading/docview/3247677686/se-2?accountid=208611
Copyright
Database copyright ProQuest LLC; ProQuest does not claim copyright in the individual underlying works.
Database
ProQuest One Academic