Multimodal Temporal Fusion Transformers Are Good Product Demand Forecasters

TL;DR

This paper introduces a multimodal transformer-based architecture that integrates visual, textual, and contextual data to improve product demand forecasting accuracy, addressing limitations of traditional methods like cold start and category dynamics.

Contribution

It proposes a novel multimodal demand forecasting model combining convolutional, graph-based, and transformer architectures, outperforming traditional approaches on real-world data.

Findings

Enhanced demand prediction accuracy across diverse products

Effective handling of cold start and category dynamics issues

Demonstrated superiority over traditional methods on large datasets

Abstract

Multimodal demand forecasting aims at predicting product demand utilizing visual, textual, and contextual information. This paper proposes a method for multimodal product demand forecasting using convolutional, graph-based, and transformer-based architectures. Traditional approaches to demand forecasting rely on historical demand, product categories, and additional contextual information such as seasonality and events. However, these approaches have several shortcomings, such as the cold start problem making it difficult to predict product demand until sufficient historical data is available for a particular product, and their inability to properly deal with category dynamics. By incorporating multimodal information, such as product images and textual descriptions, our architecture aims to address the shortcomings of traditional approaches and outperform them. The experiments conducted on a large real-world dataset show that the proposed approach effectively predicts demand for a wide range of products. The multimodal pipeline presented in this work enhances the accuracy and reliability of the predictions, demonstrating the potential of leveraging multimodal information in product demand forecasting.