LeForecast: Enterprise Hybrid Forecast by Time Series Intelligence

2025-03-27

AI Summary

To address weak model generalizability, high development/maintenance costs, and insufficient integration of heterogeneous business semantics in enterprise multi-scenario time-series forecasting (e.g., demand forecasting, inventory optimization, product planning), this paper proposes a hybrid time-series forecasting platform. Methodologically, it introduces: (1) a novel “three-pillar” modeling engine comprising a lightweight time-series foundation model (Le-TSFM), a multimodal semantic fusion module, and a plug-and-play hybrid model pool; and (2) a router-based dynamic fusion network enabling both architecture-aware and architecture-agnostic collaborative inference. Empirical evaluation across three industrial scenarios demonstrates that the platform significantly reduces redundant modeling overhead while achieving superior prediction accuracy and cross-scenario generalization compared to state-of-the-art time-series models (e.g., PatchTST, TiDE), thereby effectively supporting enterprise-scale intelligent decision-making.

Technology Category

• Machine Learning: Time-Series/Data Streams
• Planning, Routing, and Scheduling: Planning with Language Models
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Demand is spiking in industrial fields for multidisciplinary forecasting, where a broad spectrum of sectors needs planning and forecasts to streamline intelligent business management, such as demand forecasting, product planning, inventory optimization, etc. Specifically, these tasks expecting intelligent approaches to learn from sequentially collected historical data and then foresee most possible trend, i.e. time series forecasting. Challenge of it lies in interpreting complex business contexts and the efficiency and generalisation of modelling. With aspirations of pre-trained foundational models for such purpose, given their remarkable success of large foundation model across legions of tasks, we disseminate leforecast{}, an enterprise intelligence platform tailored for time series tasks. It integrates advanced interpretations of time series data and multi-source information, and a three-pillar modelling engine combining a large foundation model (Le-TSFM), multimodal model and hybrid model to derive insights, predict or infer futures, and then drive optimisation across multiple sectors in enterprise operations. The framework is composed by a model pool, model profiling module, and two different fusion approaches regarding original model architectures. Experimental results verify the efficiency of our trail fusion concepts: router-based fusion network and coordination of large and small models, resulting in high costs for redundant development and maintenance of models. This work reviews deployment of LeForecast and its performance in three industrial use cases. Our comprehensive experiments indicate that LeForecast is a profound and practical platform for efficient and competitive performance. And we do hope that this work can enlighten the research and grounding of time series techniques in accelerating enterprise.

Problem

• Develop hybrid forecasting for enterprise time series tasks
• Integrate multi-source data with advanced modeling techniques
• Optimize business operations through intelligent trend prediction

Innovation

• Integrates advanced time series data interpretations
• Combines large foundation model with multimodal hybrid
• Uses router-based fusion and model coordination