# LSTM-MSNet: Leveraging Forecasts on Sets of Related Time Series with Multiple Seasonal Patterns

arXiv: https://arxiv.org/abs/1909.04293 · HTML: https://mlsanity.com/paper/1909.04293/

## Key facts

- Cited by 23 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 10 Sep 2019.

## Built on by (most cited first, 23 of 23)

- [Monash Time Series Forecasting Archive](https://mlsanity.com/paper/2105.06643/) (2021)
- [Self-Supervised Time Series Representation Learning via Cross Reconstruction Transformer](https://mlsanity.com/paper/2205.09928/) (2022)
- [Temporal Latent Auto-Encoder: A Method for Probabilistic Multivariate Time Series Forecasting](https://mlsanity.com/paper/2101.10460/) (2021)
- [Dynamic Causal Explanation Based Diffusion-Variational Graph Neural Network for Spatio-temporal Forecasting](https://mlsanity.com/paper/2305.09703/) (2023)
- [Variance Reduced Training with Stratified Sampling for Forecasting Models](https://mlsanity.com/paper/2103.02062/) (2021)
- [Deep Learning for Time Series Forecasting: A Survey](https://mlsanity.com/paper/2503.10198/) (2025)
- [Distributional Drift Adaptation with Temporal Conditional Variational Autoencoder for Multivariate Time Series Forecasting](https://mlsanity.com/paper/2209.00654/) (2022)
- [HERMES: Hybrid Error-corrector Model with inclusion of External Signals for nonstationary fashion time series](https://mlsanity.com/paper/2202.03224/) (2022)
- [Randomized Neural Networks for Forecasting Time Series with Multiple Seasonality](https://mlsanity.com/paper/2107.01705/) (2021)
- [LoMEF: A Framework to Produce Local Explanations for Global Model Time Series Forecasts](https://mlsanity.com/paper/2111.07001/) (2021)
- [Towards Better Long-range Time Series Forecasting using Generative Forecasting](https://mlsanity.com/paper/2212.06142/) (2022)
- [Time Series Data Augmentation as an Imbalanced Learning Problem](https://mlsanity.com/paper/2404.18537/) (2024)
- [DAM: A Universal Dual Attention Mechanism for Multimodal Timeseries Cryptocurrency Trend Forecasting](https://mlsanity.com/paper/2405.00522/) (2024)
- [Lag Selection for Univariate Time Series Forecasting using Deep Learning: An Empirical Study](https://mlsanity.com/paper/2405.11237/) (2024)
- [Evaluation of Missing Data Imputation for Time Series Without Ground Truth](https://mlsanity.com/paper/2503.05775/) (2025)
- [SAMSGL: Series-Aligned Multi-Scale Graph Learning for Spatio-Temporal Forecasting](https://mlsanity.com/paper/2312.02646/) (2023)
- [Towards Better Long-range Time Series Forecasting using Generative Adversarial Networks](https://mlsanity.com/paper/2110.08770/) (2021)
- [Recurrent Neural Networks for Forecasting Time Series with Multiple Seasonality: A Comparative Study](https://mlsanity.com/paper/2203.09170/) (2022)
- [Evaluating Short-Term Forecasting of Multiple Time Series in IoT Environments](https://mlsanity.com/paper/2206.07784/) (2022)
- [Time Series Forecasting via Semi-Asymmetric Convolutional Architecture with Global Atrous Sliding Window](https://mlsanity.com/paper/2301.13691/) (2023)
- [MMformer with Adaptive Transferable Attention: Advancing Multivariate Time Series Forecasting for Environmental Applications](https://mlsanity.com/paper/2504.14050/) (2025)
- [FX-DARTS: Designing Topology-unconstrained Architectures with Differentiable Architecture Search and Entropy-based Super-network Shrinking](https://mlsanity.com/paper/2504.20079/) (2025)
- [Forecasting Energy Consumption using Recurrent Neural Networks: A Comparative Analysis](https://mlsanity.com/paper/2601.17110/) (2026)

## Abstract (arXiv)

Generating forecasts for time series with multiple seasonal cycles is an important use-case for many industries nowadays. Accounting for the multi-seasonal patterns becomes necessary to generate more accurate and meaningful forecasts in these contexts. In this paper, we propose Long Short-Term Memory Multi-Seasonal Net (LSTM-MSNet), a decomposition based, unified prediction framework to forecast time series with multiple seasonal patterns. The current state of the art in this space are typically univariate methods, in which the model parameters of each time series are estimated independently. Consequently, these models are unable to include key patterns and structures that may be shared by a collection of time series. In contrast, LSTM-MSNet is a globally trained Long Short-Term Memory network (LSTM), where a single prediction model is built across all the available time series to exploit the cross series knowledge in a group of related time series. Furthermore, our methodology combines a series of state-of-the-art multiseasonal decomposition techniques to supplement the LSTM learning procedure. In our experiments, we are able to show that on datasets from disparate data sources, like e.g. the popular M4 forecasting competition, a decomposition step is beneficial, whereas in the common real-world situation of homogeneous series from a single application, exogenous seasonal variables or no seasonal preprocessing at all are better choices. All options are readily included in the framework and allow us to achieve competitive results for both cases, outperforming many state-of-the-art multi-seasonal forecasting methods.