Salvato in:
Dettagli Bibliografici
Autori principali: Jia, Renjun, Liu, Zian, Zhu, Peng, Cheng, Dawei, Liang, Yuqi
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:https://arxiv.org/abs/2505.02880
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866911383362207744
author Jia, Renjun
Liu, Zian
Zhu, Peng
Cheng, Dawei
Liang, Yuqi
author_facet Jia, Renjun
Liu, Zian
Zhu, Peng
Cheng, Dawei
Liang, Yuqi
contents 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/}
format Preprint
id arxiv_https___arxiv_org_abs_2505_02880
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Beyond Fixed Patches: Enhancing GPTs for Financial Prediction with Adaptive Segmentation and Learnable Wavelets
Jia, Renjun
Liu, Zian
Zhu, Peng
Cheng, Dawei
Liang, Yuqi
Machine Learning
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/}
title Beyond Fixed Patches: Enhancing GPTs for Financial Prediction with Adaptive Segmentation and Learnable Wavelets
topic Machine Learning
url https://arxiv.org/abs/2505.02880