# Differential Machine Learning for Time Series Prediction

arXiv: https://arxiv.org/abs/2503.03302 · HTML: https://mlsanity.com/paper/2503.03302/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 3 of its 27 references that are in the map.
- First posted on arXiv 5 Mar 2025.

## Builds on (most cited first, 3 of 3)

- [Evaluation of deep learning models for multi-step ahead time series prediction](https://mlsanity.com/paper/2103.14250/) (2021)
- [Differential Machine Learning](https://mlsanity.com/paper/2005.02347/) (2020)
- [Mathematics of Differential Machine Learning in Derivative Pricing and Hedging](https://mlsanity.com/paper/2405.01233/) (2024)

## Abstract (arXiv)

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 R"ossler 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.