Higher Order Transformers: Enhancing Stock Movement Prediction On Multimodal Time-Series Data

TL;DR

This paper introduces Higher Order Transformers, a novel architecture that extends self-attention to better capture complex dynamics in multimodal time-series data for stock movement prediction, combining technical and fundamental signals.

Contribution

We propose Higher Order Transformers with tensor decomposition and kernel attention to improve stock prediction from multimodal data, advancing transformer capabilities in finance.

Findings

Effective in capturing complex market dynamics

Reduces computational complexity with tensor methods

Improves prediction accuracy on Stocknet dataset

Abstract

In this paper, we tackle the challenge of predicting stock movements in financial markets by introducing Higher Order Transformers, a novel architecture designed for processing multivariate time-series data. We extend the self-attention mechanism and the transformer architecture to a higher order, effectively capturing complex market dynamics across time and variables. To manage computational complexity, we propose a low-rank approximation of the potentially large attention tensor using tensor decomposition and employ kernel attention, reducing complexity to linear with respect to the data size. Additionally, we present an encoder-decoder model that integrates technical and fundamental analysis, utilizing multimodal signals from historical prices and related tweets. Our experiments on the Stocknet dataset demonstrate the effectiveness of our method, highlighting its potential for enhancing stock movement prediction in financial markets.