Demand Forecast Model Research Articles

Evaluating the Effectiveness of Time Series Transformers for Demand Forecasting in Retail

This study investigates the effectiveness of Transformer-based models for retail demand forecasting. We evaluated vanilla Transformer, Informer, Autoformer, PatchTST, and temporal fusion Transformer (TFT) against traditional baselines like AutoARIMA and AutoETS. Model performance was assessed using mean absolute scaled error (MASE) and weighted quantile loss (WQL). The M5 competition dataset, comprising 30,490 time series from 10 stores, served as the evaluation benchmark. The results demonstrate that Transformer-based models significantly outperform traditional baselines, with Transformer, Informer, and TFT leading the performance metrics. These models achieved MASE improvements of 26% to 29% and WQL reductions of up to 34% compared to the seasonal Naïve method, particularly excelling in short-term forecasts. While Autoformer and PatchTST also surpassed traditional methods, their performance was slightly lower, indicating the potential for further tuning. Additionally, this study highlights a trade-off between model complexity and computational efficiency, with Transformer models, though computationally intensive, offering superior forecasting accuracy compared to the significantly slower traditional models like AutoARIMA. These findings underscore the potential of Transformer-based approaches for enhancing retail demand forecasting, provided the computational demands are managed effectively.

Construction of e-commerce cloud-based logistics service platform by communication technology

This work aims to address the evolving demands of logistics development by proposing an innovative solution: the Intelligent Cloud-based Logistics Service Platform (LSP), which seamlessly integrates Cloud Computing (CC) and the Internet of Things (IoT). The primary objective is to enhance the efficiency and effectiveness of logistics operations through advanced technology integration. Then, short-term logistics Demand Forecasting Model (DFM) and real-time Information Tracking System (ITS) are designed based on the proposed Cloud-based LSP. Specifically, based on Deep Learning, Ensemble Empirical Mode Decomposition (EEMD), and Local Mean Decomposition (LMD), the EEMD-LMD is employed for the logistics DFM. Simultaneously, the proposed real-time logistics ITS is optimized by updating its hardware equipment through the wireless sensor. Then, the Kalman filter is employed for data processing. This work contributes to the ongoing transformation of logistics management, offering practical solutions to meet the dynamic challenges of modern supply chain management.

Judgmental adjustment of demand forecasting models using social media data and sentiment analysis within industry 5.0 ecosystems

Industry 5.0 ecosystems focus on a human-centric approach to operations and supply chain management by integrating stakeholders, advanced technologies, and processes. While incorporating social media (SM) information into demand forecasting can significantly improve accuracy, it also brings about several challenges. This paper proposes an approach to leverage Big Data originating from SM networks combined with human judgment to build demand forecasts for new products. The structured methodology is demonstrated to improve forecast accuracy in a real case of a F&B company while providing several insights into the challenges and opportunities of integrating advanced information technology into the demand forecasting process. The main challenges include effectively categorising the impact factors of SM on demand forecasting, translating insights from SM into actionable decisions, and ensuring the accuracy and reliability of the data obtained from SM networks. Future studies should involve collaborative expert input and validating the approach across various companies and industries.

Efficiency-oriented vehicle relocation of shared autonomous electric fleet in station-based car-sharing system

Long waiting delays for users and significant imbalances in vehicle distribution are bothering traditional station-based one-way electric car-sharing system operators. To address the problems above, a “demand forecast-station status judgement-vehicle relocation” multistage dynamic relocation algorithm based on the automatic formation cruising technology was proposed in this study. In stage one, a novel trip demand forecast model based on the long short-term memory network was established to predict users' car-pickup and car-return order volumes at each station. In stage two, a dynamic threshold interval was determined by combining the forecast results with the actual vehicle distribution among stations to evaluate the status of each station. Then vehicle-surplus, vehicle-insufficient, vehicle-normal stations, and the number of surplus or insufficient vehicles for each station were counted. In stage three, setting driving mileage and carbon emission as the optimization objectives, an integer linear programming mathematical model was constructed and the optimal vehicle relocation scheme was obtained by the commercial solver Gurobi. Setting 43 stations and 187 vehicles in Jiading District, Shanghai, China, as a case study, results showed that rapid vehicle rebalancing among stations with minimum carbon emissions could be realized within 15 min and the users’ car-pickup and car-return demands could be fully satisfied without any refusal.

A Water Demand Forecasting Model Based on Generative Adversarial Networks and Multivariate Feature Fusion

Accurate long-term water demand forecasting is beneficial to the sustainable development and management of cities. However, the randomness and nonlinear nature of water demand bring great challenges to accurate long-term water demand forecasting. For accurate long-term water demand forecasting, the models currently in use demand the input of extensive datasets, leading to increased costs for data gathering and higher barriers to entry for predictive projects. This situation underscores the pressing need for an effective forecasting method that can operate with a smaller dataset, making long-term water demand predictions more feasible and economically sensible. This study proposes a framework to delineate and analyze long-term water demand patterns. A forecasting model based on generative adversarial networks and multivariate feature fusion (the water demand forecast-mixer, WDF-mixer) is designed to generate synthetic data, and a gradient constraint is introduced to overcome the problem of overfitting. A multi-feature fusion method based on temporal and channel features is then derived, where a multi-layer perceptron is used to capture temporal dependencies and non-negative matrix decomposition is applied to obtain channel dependencies. After that, an attention layer receives all those features associated with the water demand forecasting, guiding the model to focus on important features and representing correlations across them. Finally, a fully connected network is constructed to improve the modeling efficiency and output the forecasting results. This approach was applied to real-world datasets. Our experimental results on four water demand datasets show that the proposed WDF-mixer model can achieve high forecasting accuracy and robustness. In comparison to the suboptimal models, the method introduced in this study demonstrated a notable enhancement, with a 62.61% reduction in the MSE, a 46.85% decrease in the MAE, and a 69.15% improve in the R2 score. This research could support decision makers in reducing uncertainty and increasing the quality of water resource planning and management.

MSR74 The Effect of Applying a Demand Forecasting Model to Assess the Accuracy of Inventory Management in a Specialty Pharmacy

MSR74 The Effect of Applying a Demand Forecasting Model to Assess the Accuracy of Inventory Management in a Specialty Pharmacy

Forecast combination with multivariate grey prediction for tourism demand forecasting

Empirical evidence has shown that forecast combination can improve the prediction accuracy of tourism demand forecasting. This paper aimed to develop a more accurate grey forecast combination method (GFCM) with multivariate grey prediction models In light of the practical applicability of grey prediction, which is not required to apply any statistical test to examine data series this research features the use of multivariate grey models through the genetic algorithm to synthesize forecasts from univariate grey prediction models commonly used in tourism forecasting into composite forecasts Empirical results showed that the proposed GFCM significantly outperformed the other combination methods considered. The results also suggested that the risk of forecast failures caused by selecting an inappropriate single model for tourism demand forecasting can be reduced by using the GFCM.

Electricity Demand, Forecasting the Peaks: Development and Implementation of C-EVA Method

Price spikes in electricity markets are very frequent, posing tremendous burden on household income and on manufacturing cost. Electricity demand (load) can be divided in two parts, energy (MWh) and peak (MW) and most of time peak is responsible for the price spikes. Literature review while devoting most of the discussion to energy, lags in the investigation of peak. In this research, a model for peak demand analysis and forecasting is developed. The model is based on a portfolio of cluster and extreme value analysis (C-EVA) methods using unit invariant knee, extremum distance estimator, and weighted scale load innovations for the optimal determination of clusters and the daily peaks divulgence. The C-EVA method consists of the Clustering part for optimal number of clusters determination and classification of day and month of peak, and the part of Extreme Value Analysis for computation of the statistical confidence interval for the load maxima. C-EVA after using all the currently available load maxima, estimates statistically the expected worst-case scenario for peaks of loads. Load peaks will be determined by EVA based on an estimated bimodal distribution while a signaling method will prompt the probability of extremes. The added value of the proposed method is that does not reject the extreme values as most methodologies do. Extreme Value Analysis for maxima and minima provide estimators for highest and lowest expected hourly load, while giving the confidence interval of the return level using an optimization method for the selection of a rolling time window, as the return period. It was found that distributed generation of renewables create a camel effect on the load peaks which increases sharpness. The proposed methodology solved this issue while opening the ground for future research for the role of storage, batteries as well as for virtual power plants as an integrated portfolio of renewables generation.

E-Commerce Demand Forecasting Using SARIMA Model and K-means Clustering Analysis

E-commerce has become a significant driver in the retail industry within the digital era, with escalating online transaction volumes. Accurate demand forecasting and effective inventory management are crucial for retailers to minimize costs related to overstock and stockouts while maximizing customer satisfaction and revenue. This study explores the application of advanced time series models for demand forecasting by e-commerce retailers, with a particular emphasis on the Seasonal Autoregressive Integrated Moving Average (SARIMA) model. The SARIMA model, an extension of the ARIMA model, incorporates autoregression, moving averages, and accounts for seasonal variations, making it suitable for analyzing data with regular periodic patterns, such as seasonal fluctuations in consumer purchasing behavior. The study methodology includes data preprocessing, establishment of the ARIMA model, construction of a Linear Regression (LR) model, development of the SARIMA model, demand forecasting, and clustering analysis. The SARIMA model demonstrated higher fit and predictive accuracy compared to traditional LR and ARIMA models, as substantiated by the 1-mWAPE and RMSE metrics. The application of the K-means clustering algorithm enhanced the homogeneity of demand forecasting and improved the model's adaptability to new sequence data. Cluster analysis categorized products into four classes, with more similar demand characteristics within each class, aiding the model in more accurately capturing potential demand changes. In conclusion, this research confirms the effectiveness and practicality of combining the SARIMA model with the K-means clustering algorithm for e-commerce demand forecasting.

Advancing urban water-energy demand predictions with a rotor hopfield neural network model optimized by contracted thermal exchange optimizer

ABSTRACT The relationship between urban water and energy demand is crucial for resource efficiency, sustainability, and environmental conservation. Rapid urbanization, population growth, and climate change necessitate integrated models that capture the complex interdependencies, feedback loops, and trade-offs between water and energy systems. This research addresses this intricate relationship by developing a modified Rotor Hopfield Neural Network (RHNN) Model using input indicators like population data, GDP, water consumption, precipitation, electricity consumption, wastewater discharge, and industrial coal usage. The model is optimized using a modified metaheuristic, called Contracted Thermal Exchange Optimizer (CTEO), resulting in a comprehensive forecast of urban water-energy demand. The model’s superior efficiency is demonstrated by comparing it with other contemporary methods. Upon comparison with alternative approaches, it is clear that the RHNN/CTEO model surpasses them, showcasing a mean relative error of 1.47% for water usage and 2.60% for energy consumption. This leads to an overall average MRE of 2.03%. This research contributes to the existing body of knowledge by offering an advanced model for urban water-energy demand forecasting, providing valuable insights for policymakers, urban planners, and stakeholders in making informed decisions related to resource allocation, infrastructure development, and sustainable urban development.

Medicine demand forecasting: an assessment of a private hospital in Pernambuco

The efficient management of materials in the healthcare sector is crucial to avoid interruptions in the treatment of hospitalized patients, especially when demand is unpredictable and based on criteria of criticality, urgency and clinical status. In complex hospital environments with high-cost materials, demand forecasting becomes essential. This study aimed to compare demand forecast models for medicines used in the urgency and emergency sector of a private hospital in the Agreste Pernambucano. The methodology involves the selection of items and data collection using the company's information system. The ABC analysis identified 27 highly relevant drugs, and different models were tested, including experience-based parameters and hyperparameter optimization. The forecasts covered the period from January to November 2023. The results indicated the Holt-Winters Additive model as most suitable for 21 medications, Holt-Winters Multiplicative for 3, and ARIMA for the others, standing out for its precision. This study strengthens decision-making in the management of medication stocks, improving demand management and ensuring continuous treatments for patients.

THEORETICAL FOUNDATIONS OF RISK MANAGEMENT FOR ENERGY SECTOR ENTERPRISES

In a rapidly evolving energy environment, effective risk management is paramount to ensuring the stability and success of businesses. This article examines a systematic approach to risk management in the energy sector, emphasizing the development of strategies adapted to a specific enterprise.The study delves into the various models used in the energy sector to manage risk. Energy balance models, energy network models, economic-mathematical models, demand and production forecasting models, risk management system models, renewable energy sources (RES) models, and system dynamic models are discussed. The study proposes a comprehensive risk management system for the energy sector, including SWOT analysis, Delphi method, sensitivity analysis and simulation modeling. The approach involves a detailed analysis of internal and external factors, expert assessments and scenario testing for the formulation of adaptive risk management strategies, taking into account regional characteristics, prioritizing resistance to change, aligning them with modern industry challenges. Given the competitive and dynamic nature of the energy sector, the study uses SWOT analysis and the Delphi method as key methodologies. The results of the study emphasize the need for a comprehensive risk management system in energy. A systematic approach combining various analytical methods is proposed, aimed at maximizing sustainability and competitiveness. The decision support system, which can be built on the basis of the developed methodology, will be automatically adapted to specific enterprises, and will facilitate prompt response to changes and optimize risk management strategies. The developed concept of a systemic approach to consideration and response to risks in the energy sector offers a clear path for enterprise risk management. The integrated system proved to be innovative and effective, making a significant contribution to the improvement of risk management methodology in the energy sector. This research provides valuable information for practical applications in energy and risk management.

Analysis of carbon neutrality technology path selection in the steel industry under policy incentives

As an energy-intensive industry, the steel industry incentivizes carbon neutrality technology innovation through carbon neutrality policies, thereby promoting the industry's low-carbon transformation. However, how policy promotes technological innovation in the steel industry to achieve net zero emissions is an urgent practical and theoretical problem that needs to be solved. This study constructs a comprehensive model that combines steel demand forecasting, technology evaluation, and bottom-up technology selection models. Described the cost-effectiveness of emission reduction technologies from a technical and economic perspective, quantitatively analyzed the emission reduction potential and evolution path of carbon neutrality technology under policy incentives in the steel industry, and provided policy support and specific deployment time for different technologies. The results indicate that under the scenario of strong subsidies and carbon prices, by 2060, the share of traditional long-process steelmaking was only 25 %, and the CO2 emissions per unit of crude steel were 0.88t CO2/t. At the same time, the implementation of policies has increased the cumulative emission reduction of the steel industry by 1.56 times from 2022 to 2060.

The Use of Material and Supply Chain Management and Service Organization to Reduce Inventory Levels While Providing Adequate Service to Customers

In the dynamic landscape of service organizations, effective Material and Supply Chain Management (MSCM) plays a pivotal role in achieving operational efficiency, cost reduction, and superior customer service. This research paper delves into the strategic application of MSCM practices to streamline inventory management processes within service-oriented industries, emphasizing the delicate balance between minimizing stock levels and meeting customer service expectations. The research aims to explore the various methodologies and technologies employed in service organizations to enhance material and supply chain operations. Emphasis will be placed on understanding how advanced technologies, data analytics, and automation contribute to minimizing excess inventory while concurrently ensuring prompt and satisfactory customer service. Key areas of investigation include the adoption of just-in-time inventory systems, demand forecasting models, and supplier relationship management strategies. The paper will critically assess case studies and industry best practices to draw insights into successful MSCM implementations and their impact on inventory levels, service quality, and overall organizational performance. Furthermore, the study will address challenges associated with implementing MSCM in service organizations, such as the inherent variability in service demand, the importance of real-time data integration, and the role of collaboration between various stakeholders within the supply chain. The anticipated outcomes of this research include providing service organizations with a comprehensive understanding of the benefits, challenges, and strategic considerations associated with adopting robust MSCM practices. The findings aim to contribute valuable insights to both academia and industry professionals seeking to optimize their material and supply chain processes for improved inventory management and enhanced customer service delivery. International Journal of Scientific Research in Engineering and Management (IJSREM) Volume: 08 Issue: 01 | January - 2024 SJIF Rating: 8.176 ISSN: 2582-3930 © 2024, IJSREM | www.ijsrem.com DOI: 10.55041/IJSREM28288 | Page 2 Keywords:Material and Supply Chain Management (MSCM),Service organizations,Inventory optimization,Customer service,Just-in-time inventory,Demand forecasting,Supplier relationship management

Improving urban water demand forecast using conformal prediction-based hybrid machine learning models

This paper presents a probabilistic forecasting method that predicts the future water consumption patterns, taking into account the inherent uncertainty in the system. The proposed method leverages statistical techniques to estimate the probability distribution of future water demand scenarios. The findings of this study will provide decision-makers with a range of possible alternatives that facilitate better planning and management of water resources. We developed a conformal prediction-based hybrid demand forecast model, which combines convolutional neural network (CNN) and bidirectional long short-term memory (BiLSTM) while comparing it with other machine learning models for probabilistic hourly water demand forecasting. Additionally, we address crucial considerations when implementing a probabilistic forecasting system, including selecting appropriate data and choosing model parameters. The performance of the proposed model is validated for probabilistic water demand forecasting in real-world settings. Results indicate noteworthy improvement in deterministic and probabilistic predictions by 10 % and 26.7 %, respectively. The findings underscore the potential benefits of this approach for improved decision-making and resource management. The results also demonstrate that the proposed hybrid model outperforms the compared machine learning models with better prediction performance and increased robustness for handling prediction uncertainties.

An integrated approach for modern supply chain management: Utilizing advanced machine learning models for sentiment analysis, demand forecasting, and probabilistic price prediction

In the contemporary business landscape, effective interpretation of customer sentiment, accurate demand forecasting, and precise price prediction are pivotal in making strategic decisions and efficiently allocating resources. Harnessing the vast array of data available from social media and online platforms, this paper presents an integrative approach employing machine learning, deep learning, and probabilistic models. Our methodology leverages the BERT transformer model for customer sentiment analysis, the Gated Recurrent Unit (GRU) model for demand forecasting, and the Bayesian Network for price prediction. These state-of-the-art techniques are adept at managing large-scale, high-dimensional data and uncovering hidden patterns, surpassing traditional statistical methods in performance. By bridging these diverse models, we aim to furnish businesses with a comprehensive understanding of their customer base and market dynamics, thus equipping them with insights to make informed decisions, optimize pricing strategies, and manage supply chain uncertainties effectively. The results demonstrate the strengths and areas for improvement of each model, ultimately presenting a robust and holistic approach to tackling the complex challenges of modern supply chain management.

Identifying potentials for Artificial Intelligence-based process support along the emergency department care pathway to alleviate overcrowding

This study examines the challenge of overcrowding within Emergency Department (ED) processes and elucidates the potential for AI-based interventions to enhance patient care and operational efficiency. We used qualitative interviews conducted in two German hospitals to identify five challenges along the ED care pathway: limited demand predictability, task overload, information unavailability, lack of IT systems interoperability, and low process standardization. We propose AI-based solutions to target these issues. For example, demand forecasting models could optimize resource allocation during patient arrival, while AI-guided queries and Decision Support Systems could improve data quality and standardization in registration & triage and diagnosis & treatment, respectively. Additionally, AI-driven text recognition and monitoring could streamline information management and patient observation. While the study is constrained by its geographic and methodological scope, it is foundational work for future research, including Design Science Research approaches to validate and implement the proposed AI-based process aids in the ED.

Demand Forecasting for Steel Company Spare Items with Time Series Templates

The rapid globalization of the world economy has required companies to become increasingly competitive. One of the factors that contribute to reducing operational costs and, consequently, helps to improve company performance is inventory management and is mainly based on demand forecasting. This represents a key factor in the management of different areas of organizations, being fundamental in managing the logistics chain, especially in inventory management, customer service levels and purchasing planning. This research project aims to test the applicability of the ARIMA, exponential smoothing and Holt-winters models for forecasting demand for spare items in a company in the steel sector. The data used in this work were extracted from the SAP material management module software. The total base available was 327.924 registered items and 42.548 were analyzed as they had a consumption history from Jan/2012 to Dec/2022. A process of classifying the consumption history was carried out as Regular, erratic, intermittent and irregular. To run the prediction models, the RStudio software was used. To measure the performance of the prediction models, the MAE metric was used. The objective of the work was achieved and it was identified that the best demand forecast model was the Hot winter with trend and seasonality model.

Time Series and Machine Learning Hybrid Models for Food Condiment Demand Forecasting: A Case Study in Thailand

Food industry is one of the most important industries in Thailand. The case-study company is a condiment manufacturer that needs to efficiently manage and plan for their business. One of the most important issues is demand forecasting. The company should precisely forecast their product demands, which will be used for operation planning. This study proposes forecasting models for both short-term and long-term planning for the company’s main condiment products. The proposed models are time series, machine learning and hybrid forecasting models which will be compared with pure time series and machine learning methods. Unlike previous work, this study proposes an innovative hybrid model, i.e., Holt- Winters exponential smoothing and Seasonal Autoregressive Integrated Moving average hybrid with Artificial neural network, which has never been considered previously. The accuracy is measured by mean absolute percentage error (MAPE) where the results are also compared to the method currently used in the company. The results show hybrid forecasting model provides the lowest overall error for both short-term and long-term forecast. The most accurate model from this paper can provide MAPE of 2.07% from short-term forecast and MAPE of 2.20% for long-term forecast (6 months in advance). When comparing with the company’s existing MAPE of 20.05%, the proposed model can increase forecast accuracy effectively.

A neural network-based model for cross-border e-commerce supply chain demand forecasting and inventory optimization

Abstract The development of the Internet makes the e-commerce transaction scale in the total global trade share grow year by year, and cross-border e-commerce has become an important growth point of global trade by virtue of its unique advantages. In this paper, the ARIMA model is used to obtain the time series demand change of a cross-border e-commerce supply chain, and the results are input into the LSTM model to realize the construction of a cross-border e-commerce supply chain demand forecasting model. The ABC inventory classification method and economic lot ordering model are used as the basis for the establishment of cross-border e-commerce supply chain inventory control strategies and multi-cycle inventory control models. Taking the sales data of WT enterprise from May 2022 to May 2023 as an example, the effectiveness of the ARIMA-LSTM model in cross-border e-commerce supply chain demand forecasting is analyzed, and the inventory control optimization results of the multi-period inventory control model are verified. The relative error fluctuation range of supply chain demand forecasting of the ARIMA-LSTM model is between [-0.1,0.2], and the cross-border e-commerce supply chain’s monthly demand forecast MAPE value is only 0.0135. After using the inventory control optimization model, the annual average inventory is reduced by 178.42 tons, and the total cost of inventory is reduced by 0.09*108 yuan. Relying on neural networks can achieve accurate prediction of cross-border e-commerce supply chain demand and optimize cross-border e-commerce supply chain inventory.