Article: Analysing trends in trading behaviour in financial markets using deep learning algorithms Journal: International Journal of Information and Communication Technology (IJICT) 2024 Vol.25 No.10 pp.34 - 46 Abstract: The complexity of financial markets demands analytical methods that capture nonlinear and time-varying data characteristics. Traditional methods often fall short, prompting the use of deep learning, particularly LSTM, for its time series processing prowess. However, LSTMs struggle with long sequences due to vanishing or exploding gradients, leading to the loss of early data significance. To address this, we introduce the LSTM-AT model, which integrates LSTM with an attention mechanism to enhance its focus on key data aspects. Our model, trained on historical financial data, outperforms traditional methods in predicting market trends. Despite its high accuracy, the model faces challenges like overfitting and data labelling requirements. Future work will focus on improving interpretability and exploring its application across various financial markets. Inderscience Publishers - linking academia, business and industry through research

Title: Analysing trends in trading behaviour in financial markets using deep learning algorithms

Authors: Yiming Tan

Addresses: Shenyang City University, Shenyang, 110112, China

Abstract: The complexity of financial markets demands analytical methods that capture nonlinear and time-varying data characteristics. Traditional methods often fall short, prompting the use of deep learning, particularly LSTM, for its time series processing prowess. However, LSTMs struggle with long sequences due to vanishing or exploding gradients, leading to the loss of early data significance. To address this, we introduce the LSTM-AT model, which integrates LSTM with an attention mechanism to enhance its focus on key data aspects. Our model, trained on historical financial data, outperforms traditional methods in predicting market trends. Despite its high accuracy, the model faces challenges like overfitting and data labelling requirements. Future work will focus on improving interpretability and exploring its application across various financial markets.

Keywords: LSTM-AT; financial markets; trading behaviour; trend prediction.

DOI: 10.1504/IJICT.2024.143332

International Journal of Information and Communication Technology, 2024 Vol.25 No.10, pp.34 - 46

Received: 12 Oct 2024
Accepted: 28 Oct 2024
Published online: 13 Dec 2024 *

Title: Analysing trends in trading behaviour in financial markets using deep learning algorithms

Keep up-to-date