A. Landenberger, S. Szénási: Financial Forecasting using Quantum-Inspired Deep Learning. In IEEE 29th International Conference on Intelligent Engineering Systems (INES 2025) Proceedings. pp. 325–328, 2025. ISBN 979-8-3315-9771-9 link

Abstract: Forecasting stock market prices has always been a difficult field, particularly with the incorporation of multiple scientific fields, adding to its complexity. This study presents a deep machine learning model inspired by quantum principles that combines classical machine learning methods with quantum algorithms to improve predictive precision. The study aims to create a hybrid model by integrating historical trading data, sentiment-analyzed news, and market indicators from multiple data sources. The model design includes a transformer neural network combined with quantum-inspired layers, allowing for the effective recognition of intricate patterns. The study shows that integrating quantum-inspired methods into the transformer model can enhance prediction precision over conventional deep learning models. The uniqueness of the model lies in its ability to deliver highly precise stock closing price predictions, even with smaller datasets, achieving high accuracy, thereby offering a significant advantage in the stock market. This provides a significant competitive edge in stock market forecasting.