Forecasting of Multivariate Time Series Data of Export and Import Volumes of Coffee Using the Vector Autoregressive (VAR) - Long Short-Term Memory (LSTM) Hybrid Model

DOI

10.2991/978-94-6463-730-4_5How to use a DOI?

Keywords

Forecasting; Multivariate Time Series; Vector Autoregressive (VAR); Long Short-Term Memory (LSTM); VAR-LSTM Hybrid Model

Abstract

The primary objective of this study is to develop a hybrid forecasting model that integrates the Vector Autoregressive (VAR) model with Long Short-Term Memory (LSTM) networks to predict multivariate time series data on coffee export and import volumes. The VAR-LSTM hybrid model represents an advanced methodological approach that combines the strengths of both techniques: the VAR model’s ability to capture linear interdependencies and the LSTM network’s proficiency in modeling complex non-linear patterns. By integrating these models, this study leverages their complementary advantages to enhance forecasting accuracy for multivariate time series data. This hybrid approach is particularly beneficial as it addresses both linear and non-linear dependencies within the dataset. The results indicate that the VAR-LSTM model effectively replicates the observed data patterns, demonstrating strong predictive performance. The graphical analysis provides compelling evidence of the model’s consistency and accuracy, with a Mean Absolute Percentage Error (MAPE) of 0.41% (corresponding to an accuracy of 99.59%) for the predicted data. In comparison, the classical VAR model achieved a MAPE of 0.54% (accuracy of 99.46%). These findings highlight the reliability and accuracy of the VAR-LSTM hybrid model in improving forecasting precision for multivariate time series data.

Copyright

© 2025 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - Warsono Warsono
AU  - Dian Kurniasari
AU  - Dyah Aring Hepiana Lestari
AU  - Heru Wahyudi
AU  - Mawar Alhani
AU  - Arif Su’admamaji
AU  - Muhtarom Ahkam Maulana
PY  - 2025
DA  - 2025/05/27
TI  - Forecasting of Multivariate Time Series Data of Export and Import Volumes of Coffee Using the Vector Autoregressive (VAR) - Long Short-Term Memory (LSTM) Hybrid Model
BT  - Proceedings of the 5th International Conference on Applied Sciences, Mathematics, and Informatics (ICASMI 2024)
PB  - Atlantis Press
SP  - 43
EP  - 56
SN  - 2352-541X
UR  - https://doi.org/10.2991/978-94-6463-730-4_5
DO  - 10.2991/978-94-6463-730-4_5
ID  - Warsono2025
ER  -