Multiscale Attention-Based Neural Network for Stock and Cryptocurrency Price Prediction: Bridging Microstructure and Macro Trends

Zichao Li, Zong Ke

Keywords: Multiscale Modeling, Attention Mechanism, Financial Forecasting, Stock Market, Cryptocurrency, Algorithmic Trading, Risk Management

TL;DR: This paper introduces MANN, a Multiscale Attention-Based Neural Network, for accurate stock and cryptocurrency price prediction, outperforming traditional and deep learning models across diverse datasets and time horizons.

Confirmation Of Submission Requirements: I submit an abstract. It uses the template provided on the submission page and is no longer than 2 pages.

Submission Number: 287