A Hybrid Temporal Convolutional Network and Transformer Model for Accurate and Scalable Sales Forecasting

Accurate product sales forecasting is critical for inventory management, pricing strategies, and supply chain optimization in the retail industry. This article proposes a novel deep learning architecture that integrates Temporal Convolutional Networks (TCNs) with Transformer-based attention mechanis...

Full description

Saved in:
Bibliographic Details
Main Authors: Rafi MD AL, Gourab Nicholas Rodrigues, Nazmul Hossain Mir MD, Shahriar Mahmud Bhuiyan MD, M. F. Mridha, MD Rashedul Islam, Yutaka Watanobe
Format: Article
Language:English
Published: IEEE 2025-01-01
Series:IEEE Open Journal of the Computer Society
Subjects:
Online Access:https://ieeexplore.ieee.org/document/10870315/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849708088798478336
author Rafi MD AL
Gourab Nicholas Rodrigues
Nazmul Hossain Mir MD
Shahriar Mahmud Bhuiyan MD
M. F. Mridha
MD Rashedul Islam
Yutaka Watanobe
author_facet Rafi MD AL
Gourab Nicholas Rodrigues
Nazmul Hossain Mir MD
Shahriar Mahmud Bhuiyan MD
M. F. Mridha
MD Rashedul Islam
Yutaka Watanobe
author_sort Rafi MD AL
collection DOAJ
description Accurate product sales forecasting is critical for inventory management, pricing strategies, and supply chain optimization in the retail industry. This article proposes a novel deep learning architecture that integrates Temporal Convolutional Networks (TCNs) with Transformer-based attention mechanisms to capture both short-term and long-term dependencies in time-series sales data. Utilizing the Favorita Grocery Sales Forecasting dataset, our hybrid TCN Transformer model demonstrates superior performance over existing models by incorporating external factors such as holidays, promotions, oil prices, and transaction data. The model achieves state-of-the-art results with a Mean Absolute Error (MAE) of 2.01, Root Mean Squared Error (RMSE) of 2.81, and a Weighted Mean Absolute Percentage Error (wMAPE) of 4.22%, significantly outperforming other leading models such as LSTM, GRU, and TFT. Extensive cross-validation confirms the robustness of our model, achieving consistently high performance across multiple folds.
format Article
id doaj-art-3247ca274ad6498f9c5dde2d2cc686fc
institution DOAJ
issn 2644-1268
language English
publishDate 2025-01-01
publisher IEEE
record_format Article
series IEEE Open Journal of the Computer Society
spelling doaj-art-3247ca274ad6498f9c5dde2d2cc686fc2025-08-20T03:15:47ZengIEEEIEEE Open Journal of the Computer Society2644-12682025-01-01638039110.1109/OJCS.2025.353857910870315A Hybrid Temporal Convolutional Network and Transformer Model for Accurate and Scalable Sales ForecastingRafi MD AL0https://orcid.org/0009-0000-0115-7032Gourab Nicholas Rodrigues1https://orcid.org/0009-0000-4119-1753Nazmul Hossain Mir MD2https://orcid.org/0009-0001-7390-3046Shahriar Mahmud Bhuiyan MD3https://orcid.org/0009-0009-6126-6404M. F. Mridha4https://orcid.org/0000-0001-5738-1631MD Rashedul Islam5https://orcid.org/0000-0001-8676-6338Yutaka Watanobe6https://orcid.org/0000-0002-0030-3859Washington University of Science and Technology, Alexandria, VA, USAWashington University of Science and Technology, Alexandria, VA, USAWashington University of Science and Technology, Alexandria, VA, USAWashington University of Science and Technology, Alexandria, VA, USADepartment of Computer Science and Engineering, American International University-Bangladesh, Dhaka, BangladeshDepartment of Computer Science and Engineering, University of Asia Pacific, Dhaka, BangladeshSchool of Computer Science and Engineering, University of Aizu, Aizu-wakamatsu, JapanAccurate product sales forecasting is critical for inventory management, pricing strategies, and supply chain optimization in the retail industry. This article proposes a novel deep learning architecture that integrates Temporal Convolutional Networks (TCNs) with Transformer-based attention mechanisms to capture both short-term and long-term dependencies in time-series sales data. Utilizing the Favorita Grocery Sales Forecasting dataset, our hybrid TCN Transformer model demonstrates superior performance over existing models by incorporating external factors such as holidays, promotions, oil prices, and transaction data. The model achieves state-of-the-art results with a Mean Absolute Error (MAE) of 2.01, Root Mean Squared Error (RMSE) of 2.81, and a Weighted Mean Absolute Percentage Error (wMAPE) of 4.22%, significantly outperforming other leading models such as LSTM, GRU, and TFT. Extensive cross-validation confirms the robustness of our model, achieving consistently high performance across multiple folds.https://ieeexplore.ieee.org/document/10870315/Sales forecastingtime-series forecastingtemporal convolutional networkstransformersand deep learning
spellingShingle Rafi MD AL
Gourab Nicholas Rodrigues
Nazmul Hossain Mir MD
Shahriar Mahmud Bhuiyan MD
M. F. Mridha
MD Rashedul Islam
Yutaka Watanobe
A Hybrid Temporal Convolutional Network and Transformer Model for Accurate and Scalable Sales Forecasting
IEEE Open Journal of the Computer Society
Sales forecasting
time-series forecasting
temporal convolutional networks
transformers
and deep learning
title A Hybrid Temporal Convolutional Network and Transformer Model for Accurate and Scalable Sales Forecasting
title_full A Hybrid Temporal Convolutional Network and Transformer Model for Accurate and Scalable Sales Forecasting
title_fullStr A Hybrid Temporal Convolutional Network and Transformer Model for Accurate and Scalable Sales Forecasting
title_full_unstemmed A Hybrid Temporal Convolutional Network and Transformer Model for Accurate and Scalable Sales Forecasting
title_short A Hybrid Temporal Convolutional Network and Transformer Model for Accurate and Scalable Sales Forecasting
title_sort hybrid temporal convolutional network and transformer model for accurate and scalable sales forecasting
topic Sales forecasting
time-series forecasting
temporal convolutional networks
transformers
and deep learning
url https://ieeexplore.ieee.org/document/10870315/
work_keys_str_mv AT rafimdal ahybridtemporalconvolutionalnetworkandtransformermodelforaccurateandscalablesalesforecasting
AT gourabnicholasrodrigues ahybridtemporalconvolutionalnetworkandtransformermodelforaccurateandscalablesalesforecasting
AT nazmulhossainmirmd ahybridtemporalconvolutionalnetworkandtransformermodelforaccurateandscalablesalesforecasting
AT shahriarmahmudbhuiyanmd ahybridtemporalconvolutionalnetworkandtransformermodelforaccurateandscalablesalesforecasting
AT mfmridha ahybridtemporalconvolutionalnetworkandtransformermodelforaccurateandscalablesalesforecasting
AT mdrashedulislam ahybridtemporalconvolutionalnetworkandtransformermodelforaccurateandscalablesalesforecasting
AT yutakawatanobe ahybridtemporalconvolutionalnetworkandtransformermodelforaccurateandscalablesalesforecasting
AT rafimdal hybridtemporalconvolutionalnetworkandtransformermodelforaccurateandscalablesalesforecasting
AT gourabnicholasrodrigues hybridtemporalconvolutionalnetworkandtransformermodelforaccurateandscalablesalesforecasting
AT nazmulhossainmirmd hybridtemporalconvolutionalnetworkandtransformermodelforaccurateandscalablesalesforecasting
AT shahriarmahmudbhuiyanmd hybridtemporalconvolutionalnetworkandtransformermodelforaccurateandscalablesalesforecasting
AT mfmridha hybridtemporalconvolutionalnetworkandtransformermodelforaccurateandscalablesalesforecasting
AT mdrashedulislam hybridtemporalconvolutionalnetworkandtransformermodelforaccurateandscalablesalesforecasting
AT yutakawatanobe hybridtemporalconvolutionalnetworkandtransformermodelforaccurateandscalablesalesforecasting