Saved in:
Bibliographic Details
Main Authors: Yadav, Akash, Nualart, Eulalia
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2503.03302
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866929748335132672
author Yadav, Akash
Nualart, Eulalia
author_facet Yadav, Akash
Nualart, Eulalia
contents Accurate time series prediction is challenging due to the inherent nonlinearity and sensitivity to initial conditions. We propose a novel approach that enhances neural network predictions through differential learning, which involves training models on both the original time series and its differential series. Specifically, we develop a differential long short-term memory (Diff-LSTM) network that uses a shared LSTM cell to simultaneously process both data streams, effectively capturing intrinsic patterns and temporal dynamics. Evaluated on the Mackey-Glass, Lorenz, and Rössler chaotic time series, as well as a real-world financial dataset from ACI Worldwide Inc., our results demonstrate that the Diff- LSTM network outperforms prevalent models such as recurrent neural networks, convolutional neural networks, and bidirectional and encoder-decoder LSTM networks in both short-term and long-term predictions. This framework offers a promising solution for enhancing time series prediction, even when comprehensive knowledge of the underlying dynamics of the time series is not fully available.
format Preprint
id arxiv_https___arxiv_org_abs_2503_03302
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Differential Machine Learning for Time Series Prediction
Yadav, Akash
Nualart, Eulalia
Machine Learning
Accurate time series prediction is challenging due to the inherent nonlinearity and sensitivity to initial conditions. We propose a novel approach that enhances neural network predictions through differential learning, which involves training models on both the original time series and its differential series. Specifically, we develop a differential long short-term memory (Diff-LSTM) network that uses a shared LSTM cell to simultaneously process both data streams, effectively capturing intrinsic patterns and temporal dynamics. Evaluated on the Mackey-Glass, Lorenz, and Rössler chaotic time series, as well as a real-world financial dataset from ACI Worldwide Inc., our results demonstrate that the Diff- LSTM network outperforms prevalent models such as recurrent neural networks, convolutional neural networks, and bidirectional and encoder-decoder LSTM networks in both short-term and long-term predictions. This framework offers a promising solution for enhancing time series prediction, even when comprehensive knowledge of the underlying dynamics of the time series is not fully available.
title Differential Machine Learning for Time Series Prediction
topic Machine Learning
url https://arxiv.org/abs/2503.03302