Beyond Fixed Patches: Enhancing GPTs for Financial Prediction with Adaptive Segmentation and Learnable Wavelets

TL;DR

This paper introduces GPT4FTS, a novel framework that improves financial time series prediction by using adaptive segmentation and learnable wavelet transforms to capture multi-scale market patterns.

Contribution

It proposes a new method combining dynamic patch segmentation and learnable wavelet modules to enhance GPTs for financial forecasting.

Findings

Outperforms traditional models on real-world datasets.

Effectively captures multi-scale market patterns.

Demonstrates flexibility with dynamic time-frequency features.

Abstract

The extensive adoption of web technologies in the finance and investment sectors has led to an explosion of financial data, which contributes to the complexity of the forecasting task. Traditional machine learning models exhibit limitations in this forecasting task constrained by their restricted model capacity. Recent advances in Generative Pre-trained Transformers (GPTs), with their greatly expanded parameter spaces, demonstrate promising potential for modeling complex dependencies in temporal sequences. However, existing pretraining-based approaches typically focus on fixed-length patch analysis, ignoring market data's multi-scale pattern characteristics. In this study, we propose $\mathbf{GPT4FTS}$, a novel framework that enhances pretrained transformer capabilities for temporal sequence modeling through dynamic patch segmentation and learnable wavelet transform modules. Specifically, we first employ K-means++ clustering based on DTW distance to identify scale-invariant patterns in market data. Building upon pattern recognition results, we introduce adaptive patch segmentation that partitions temporal sequences while preserving pattern integrity. To accommodate time-varying frequency characteristics, we devise a dynamic wavelet transform module that emulates discrete wavelet transformation with enhanced flexibility in capturing time-frequency features. Extensive experiments on real-world financial datasets substantiate the framework's efficacy. The source code is available: \href{https://anonymous.4open.science/r/GPT4FTS-6BCC/}