Demand Prediction Research Articles

Research on Public Space Area Indicators of Physical Examination Centers

With the increasing demand for medical checkups, problems such as overcrowding and confusion in the public spaces of physical examination centers are becoming prominent. This seriously affects the efficiency of medical checkups and the satisfaction of patients. Because of the lack of reference norms and standards, spatial resources are allocated blindly in physical examination centers. Therefore, the reasonable prediction of demands for space and the allocation of areas for public space are problems we need to explore. This study aims to establish a foundation for spatial resource allocation by analyzing the relationship between area allocation and volume of medical checkups and spatial comfort. The objective is to balance service quality and construction costs in physical examination centers. Utilizing AnyLogic, this research simulates pedestrian flow characteristics during peak hours based on empirical data. Additionally, it provides statistics on and analysis of such areas in 30 physical examination centers. The findings reveal that the average per capita allocation of public space ranges from 6.90 to 7.10 m2. The functional relationship between the area of public space (y) and the total area (x) is y = 0.3649x − 57.096. This paper also proposes a relationship between the waiting space and the daily volume of medical checkups according to two comfort levels. The percentages of waiting space in public space ranges from 10% to 12.5% (comfort level) and 7% to 10% (general comfort level). This research provides architects with a specific and operable foundation for space planning, enhancing the standardization and scientification of physical examination center construction. Furthermore, the methodology used in this study proves valuable for future analyses of the interrelationships among functional units in medical buildings. It can also be applied during the postuse evaluation phase, enabling continuous assessment and iterative improvement in the design.

Interpretable machine learning models for displacement demand prediction in reinforced concrete buildings under pulse-like earthquakes

This work proposes a novel procedure to guide the development of machine learning models for estimating the seismic demand in existing reinforced concrete (RC) buildings. The proposed approach is organized across two scales. A large-scale (nonparametric) machine learning model is first obtained by means of Gaussian Process Regression (GPR) using all candidate building attributes and intensity measures. SHapley Additive exPlanations (SHAP) values are utilized to facilitate its interpretation and to assist the rational selection of a small subset of intensity measures, which is finally employed to develop a (symbolic) reduced-scale machine learning model by means of Genetic Programming (GP). Simplified models of archetype buildings are adopted to develop machine learning techniques at both scales, in such a way to alleviate the simulation time for preparing large datasets. Refined models representative of actual buildings are instead considered for the unbiased final assessment.The proposed approach is applied to develop predictive machine learning models for the maximum inter-storey drift in bare frames, pilotis frames and frames with infills under pulse-like seismic ground motions. Consequently, the critical examination of the SHAP values revealed the most significant intensity measures and unfolded interesting patterns depending on the occupancy rate of the infills. Moreover, the final assessment demonstrates that this approach allows the management of a non-homogeneous building stock consisting of very diverse structural systems (i.e., spanning from existing buildings designed against gravity loads only to buildings that comply with outdated seismic codes) while providing satisfactory predictions of the seismic demand with minimum computational effort.

One Method for Predicting Satellite Communication Terminal Service Demands Based on Artificial Intelligence Algorithms

This paper presents a traffic demand prediction method based on deep learning algorithms, aiming to address the dynamic traffic demands in satellite communication and enhance resource management efficiency. Integrating Seq2Seq and LSTM networks, the method improves prediction accuracy and applicability, especially for mobile terminals such as aviation and maritime ones. Unlike traditional approaches, it does not require extensive statistical data and can be generalized to real-world systems, providing stable long-term traffic demand predictions. This study utilizes real-world flight data mapped to corresponding satellite beams, allowing the precise prediction of beam-specific traffic demands. The results show that aggregating historical demand data for beams with similar trends achieves an average predictive Mean Squared Error (MSE) of 0.0007 and a maximum MSE fluctuation of 0.014, significantly outperforming predictions based on average values in terms of stability and accuracy. This novel solution for resource management in satellite communication ensures efficient and accurate long-term traffic demand predictions.

The correlation of BMI and insulin resistance in moderately burned patients.

<b>Introduction:</b> Obesity is strongly related to serious comorbidities that might affect the healing process. Elevated Body Mass Index (BMI) and insulin resistance have a significant impact on the development of the metabolic syndrome often leading to lethal cardiovascular complications.<b>Aim:</b> The aim of the study was to verify the correlation of BMI and insulin resistance with clinical parameters of moderately burned patients.<b>Materials and methods:</b> There were 149 patients enrolled in the study and their clinical data was retrospectively analyzed. The laboratory tests, insulin demand, BMI, and surgical procedures were evaluated on admission and discharge.<b>Results:</b> Burned patients who required insulin were characterized by worse laboratory results on admission to the burn unit, they had lower hemoglobin (HGB) levels (P = 0.0001), higher creatinine levels by 0.323 units (P = 0.009), higher C-reactive protein (CRP) by approximately 94 units (P = 0.0001), as higher procalcitonin (PCT) by approximately 0.5 units (P = 0.001) as compared to non-insulin-treated patients. Moreover, burned patients who required insulin stayed in the hospital for an average of 10 days longer. All patients from the insulin-demand subgroup had elevated triglycerides (Tg) levels on admission with increased indexes of insulin resistance.<b>Discussion:</b> Our study suggests that the protective effect of a higher BMI in burned patients, known as the 'obesity paradox' may be compromised by insulin resistance.<b>Conclusions and significance of the study:</b> The results show that elevated Tg on admission to the burn unit coexisting with a BMI over 25 kg/m<sup>2</sup> may be used as an important prognostic factor and may help with prediction of insulin demand and worse outcome in moderately burned patients.

A novel method for long-term power demand prediction using enhanced data decomposition and neural network with integrated uncertainty analysis: A Cuba case study

This study developed a methodological approach for long-term electricity demand forecasting and applied it to the electricity demand in Cuba, which is crucial for transitioning from a fossil fuel-dependent system to renewable energy sources. The methodology employs enhanced complete ensemble empirical mode decomposition with adaptive noise (ECEEMDAN) applied for obtaining long-term trends from historical electricity usage data decomposition, combined with a long short-term memory (LSTM) deep learning model for prediction. Comprehensive datasets, including historical electricity consumption, economic indicators, and demographic data, are utilized in the analysis. Monte Carlo simulations, then, are integrated to address uncertainties in prediction and explore 50 different scenarios of future electricity demand. The study forecasts varying scenarios for the energy demand of Cuba by 2050, with the extreme low scenario projecting a decrease of up to 7.9% compared to the 2019 level. This research offers a groundbreaking framework specifically designed to aid Cuba's energy sector stakeholders in informed decision-making during this critical energy transition. The adaptability of the methodology makes it applicable for long-term projections in various sectors, offering a reliable tool for global decision makers.

Short-term Power Load Forecasting Based on EMD-GWO-BP

Abstract The short-range prediction of electricity demand holds immense importance in the strategic planning and advancement of the energy sector. However, the actual load sequence data reflects multiple complex properties, such as nonlinearity, non-stationarity, and temporal variation. In order to accurately forecast the load, this paper presents a three-level hybrid integrated short-term load prediction method composed of empirical mode decomposition (EMD), Grey Wolf Optimization (GWO) and BP neural network (BP). EMD is used to decompose load data to obtain good power consumption characteristics; GWO is used to obtain the optimal weight and threshold required for BP prediction. The hybrid integration method (EMD-GWO-BP) was evaluated using the 2017 annual data of Chaoyang County, Liaoning Province. The EMD-GWO-BP method was compared with the other two mainstream coupling methods (BP, GWO-BP). The statistical analysis Indicates that the suggested method within this document shows better forecast precision on three standard scales of MAPE, MAD and RMSE, which reflects the advanced nature of this method.

Research on urban water demand prediction based on machine learning and feature engineering

ABSTRACT Urban water demand prediction is not only the foundation of water resource planning and management, but also an important component of water supply system optimization and scheduling. Therefore, predicting future water demand is of great significance. For univariate time series data, the issue of outliers can be solved through data preprocessing. Then, the data input dimension is increased through feature engineering, and finally, the LightGBM (Light Gradient Boosting Machine) model is used to predict future water demand. The results demonstrate that cubic polynomial interpolation outperforms the Prophet model and the linear method in the context of missing value interpolation tasks. In terms of predicting water demand, the LightGBM model demonstrates excellent forecasting performance and can effectively predict future water demand trends. The evaluation indicators MAPE (mean absolute percentage error) and NSE (Nash–Sutcliffe efficiency coefficient) on the test dataset are 4.28% and 0.94, respectively. These indicators can provide a scientific basis for short-term prediction of water supply enterprises.

Emergency Material Prediction in South China Based on Case Analysis Method

After the occurrence of flood disasters, timely and rapid emergency response is one of the important tasks to improve the efficiency of emergency rescue and reduce disaster losses, and the prediction of material demand, which is the premise and basis of emergency response in disaster areas after disasters, is one of the key problems that need to be solved urgently. Considering the characteristics of poor information in the post-disaster disaster area, a post-disaster material demand forecasting technology based on fuzzy case reasoning was introduced. Firstly, on the basis of the analysis and summary of the existing case database, the key feature attributes of flood disasters that affect the demand for materials after the disaster are extracted, and the fuzzy set of flood disaster characteristics is established by introducing the concept of fuzzy sets, and then the membership degree of the specific feature attribute values of the new and old cases to the fuzzy set is calculated, in order to measure the similarity between the new and old cases, and the corrected measure closeness based on the weight of the feature attributes of the new and old cases is calculated, and the one with the largest closeness represents the best match between the new and old cases. Finally, an actual case is used to show the specific application process of the technology, and the existing reference cases that are closest to the new case are obtained, which can provide reference for post-disaster emergency rescue. The results show that the proposed method can provide a scientific prediction of the demand for emergency supplies and provide a basis for decision-making for the relevant emergency management departments in an actual flood disaster case in South China.

Application of the K-Nearest Neighbor Method to Predict Demand for Goods from Customers at PT Sinergi Prima Enjineering

PT Sinergi Prima Enjineering, which is engaged in services, has been trusted as a contractor in several companies facing challenges in handling the large number of requests for goods and stock inventory management. This research aims to improve the prediction of demand for goods and inventory management using the calculation of the K-Nearest Neighbor (KNN) method and RapidMiner tools. With the comparison of calculations between KNN and RapidMiner using ten test data, the results are appropriate where the categorical grouping is often ordered totaling five data, moderately ordered totaling two data and rarely ordered totaling three data. The test results show that K = 3 produces a prediction accuracy of 91.98%. These results show that K-Nearest Neighbor can accurately anticipate future stock inventory and items that will be ordered by customers and it is hoped that the company can improve customer satisfaction and overall operational performance.

Health supply chain forecasting: a comparison of ARIMA and LSTM time series models for demand prediction of medicines

Health supply chain forecasting: a comparison of ARIMA and LSTM time series models for demand prediction of medicines

Evaluating and Comparing AI Models for Hourly Energy Demand Prediction

This study investigates the application of various AI models to predict energy demand, comparing the performance of four specific models: Decision Tree, Random Forest, Gradient Boosting, and Linear Regression. The evaluation of these models' prediction performance reveals that ensemble methods like Random Forest and Gradient Boosting exhibit promising generalization capabilities, while the Decision Tree model shows high training accuracy but suffers from overfitting. The discussion underscores the importance of ensemble techniques and feature engineering optimization in mitigating overfitting and enhancing forecast accuracy. Furthermore, the study highlights the potential of AI-driven approaches to promote sustainability and resilience in energy systems, emphasizing the need for further optimization and collaboration among stakeholders to achieve a cleaner, more sustainable energy future.

Federated Learning-Based Prediction of Energy Consumption from Blockchain-Based Black Box Data for Electric Vehicles

In modern society, the proliferation of electric vehicles (EVs) is continuously increasing, presenting new challenges that necessitate integration with smart grids. The operational data from electric vehicles are voluminous, and the secure storage and management of these data are crucial for the efficient operation of the power grid. This paper proposes a novel system that utilizes blockchain technology to securely store and manage the black box data of electric vehicles. By leveraging the core characteristics of blockchain—immutability and transparency—the system records the operational data of electric vehicles and uses federated learning (FL) to predict their energy consumption based on these data. This approach allows the balanced management of the power grid’s load, optimization of energy supply, and maintenance of grid stability while reducing costs. Additionally, the paper implements a searchable black box data storage system using a public blockchain, which offers cost efficiency and robust anonymity, thereby enhancing convenience for electric vehicle users and strengthening the stability of the power grid. This research presents an innovative approach to the integration of electric vehicles and smart grids, exploring ways to enhance the stability and energy efficiency of the power grid. The proposed system has been validated through real data and simulations, demonstrating its effectiveness and performance in managing black box data and predicting energy consumption, thereby improving the efficiency and stability of the power grid. This system is expected to empower electric vehicle users with data ownership and provide power suppliers with more accurate energy demand predictions, promoting sustainable energy consumption and efficient power grid operations.

Transfer learning for cross-modal demand prediction of bike-share and public transit

The urban transportation system is a combination of multiple transport modes, and the interdependencies across those modes exist. This means that the travel demand across different travel modes could be correlated as one mode may receive/create demand from/for another mode, not to mention natural correlations between different demand time series due to general demand flow patterns across the network. It is expected that cross-modal ripple effects will become more prevalent with Mobility as a Service. Therefore, by propagating demand data across modes, a better demand prediction could be obtained. To this end, this study explores various transfer learning strategies and machine learning models for cross-modal demand prediction. The trip data of bike-share, metro, and taxi are processed as the station-level passenger flows, and then the proposed prediction method is tested in the large-scale case studies of Nanjing and Chicago. The results suggest that prediction models with transfer learning perform better than unimodal prediction models. Fine-tuning without freezing strategy performs the best among all transfer learning strategies, and the split-brain strategy can handle the data missing problem. Furthermore, the 3-layer stacked Long Short-Term Memory model performs particularly well in cross-modal demand prediction. These results verify our deep transfer learning method’s forecasting improvement over existing benchmarks and demonstrate the good transferability for cross-modal demand prediction in multiple cities.

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.

Prediction of footwear demand using Prophet and SARIMA

In an increasingly globalized market, where world container traffic since 2000 has almost quadrupled, the prediction of demand is an element of great importance for the optimal business development of a company. This work focuses on demand forecasting in the fashion sector. It is a very volatile market, with some characteristics such as: seasonality, culture and fashion trends, that makes it difficult to estimate the inter-seasonal footwear demand. In recent years, many algorithms for the prediction of demand have been studied; they can be divided into three natures: statistical algorithms, artificial intelligence algorithms and hybrid algorithms, each of them with its own characteristics. AI-generated predictions provide business professionals with the ability to organize the purchase of materials, manage production processes and stock quantity. Therefore, the purpose of this work is to forecast the long-term sales of a highly seasonal footwear model based on its historical data, using the Prophet and SARIMA (Seasonal Autoregressive Integrated Moving Average) algorithms. This represents a novelty as sales predictions for footwear in the state of the art are not typically made over the long term or highly seasonal, and there is no model that clearly outperforms others. Additionally, a set of Key Performance Indicators has been established to evaluate the prediction outcomes, as the same indicators such as MAE, MAPE and RMSE are commonly used in the state of the art. Furthermore, a relational database structure has been proposed for the organized storage of future predictions. Finally, the results between Prophet and SARIMA have been compared to ascertain whether Prophet (a non-linear statistical algorithm) outperforms SARIMA (a linear statistical algorithm). In the model prediction Prophet obtains an accuracy of 98.8% and a 158.8 MAE, while SARIMA reaches an accuracy of 93% and a 83.9 MAE; all in all really positive results taking into account long-term prediction and high seasonality. It has been observed how Prophet provides better results when it comes to predicting results of annual quantities, for example, the number of shoes that are expected to be sold next year. However, SARIMA returns better results for those KPI that are calculated considering the monthly distribution of the prediction, as well as being 15 times faster in the mean prediction time.

A City Shared Bike Dispatch Approach Based on Temporal Graph Convolutional Network and Genetic Algorithm.

Public transportation scheduling aims to optimize the allocation of resources, enhance efficiency, and increase passenger satisfaction, all of which are crucial for building a sustainable urban transportation system. As a complement to public transportation, bike-sharing systems provide users with a solution for the last mile of travel, compensating for the lack of flexibility in public transportation and helping to improve its utilization rate. Due to the characteristics of shared bikes, including peak usage periods in the morning and evening and significant demand fluctuations across different areas, optimizing shared bike dispatch can better meet user needs, reduce vehicle vacancy rates, and increase operating revenue. To address this issue, this article proposes a comprehensive decision-making approach for spatiotemporal demand prediction and bike dispatch optimization. For demand prediction, we design a T-GCN (Temporal Graph Convolutional Network)-based bike demand prediction model. In terms of dispatch optimization, we consider factors such as dispatch capacity, distance restrictions, and dispatch costs, and design an optimization solution based on genetic algorithms. Finally, we validate the approach using shared bike operating data and show that the T-GCN can effectively predict the short-term demand for shared bikes. Meanwhile, the optimization model based on genetic algorithms provides a complete dispatch solution, verifying the model's effectiveness. The shared bike dispatch approach proposed in this paper combines demand prediction with resource scheduling. This scheme can also be extended to other transportation scheduling problems with uncertain demand, such as store replenishment delivery and intercity inventory dispatch.

Application of Improved Genetic Algorithm and Deep Learning in Cold Chain Logistics Distribution Demand Prediction

In order to solve the problem of inaccurate prediction results caused by the excessive impact of downstream on upstream suppliers in the process of cold chain logistics transportation demand prediction, the author proposes a demand prediction system Multi agent based on improved genetic algorithm and deep learning. The system will improve genetic algorithm, combine deep learning with practical problems in cold chain logistics supply chain, and evaluate the improved model through instance simulation. The results are as follows: After optimization, the order quantity of each stratum reduces the influence of retailers on upstream suppliers by more than 70%. As for the overall transportation cost, it shows a continuous upward trend within 20 cycles, while the total cost of each cycle fluctuates in a lower range after optimization, reducing the overall total cost by about 50%. It shows the reliability of the improved demand forecasting system in this study to greatly reduce storage and transportation costs.

Application of ANN in Tourism Business Development for Demand Forecasting and Management of Tourism Headcount

With the rapid development of tourism, it is imperative to forecast tourism demand to maintain the long-term stable development of the tourism industry and make good planning for future tourism enterprises. The study uses the classical model of artificial neural network-BP neural network for tourism number demand prediction, given the problems of traditional BP neural networks, such as prematurity and poor convergence speed, this paper studies the iterative optimization of the algorithm of particle swarm fusion immune mechanism and finds out the optimal network parameters, to build an IAPSO-BP tourism demand prediction model. Tourist amounts from 2007 to 2017 in certain area-related data samples, the training model of iterative speed and fitting effect, and the rolling forecasting method will be used to predict the 2018-2022 years of travel. It can be seen from the convergence curve that the convergence speed of parameter optimization of the IAPSO algorithm is the fastest; the improvedIAPSO-BP network has the best training fitting effect, with a relative average error of 2.03% and an absolute average error of 4.37%, which is better than other forecasting methods. The IAPSO-BP prediction model has higher accuracy and better performance, which can provide an effective basis for the development planning of tourism enterprises and has higher practical application value.

Research on Product Demand Forecasting and Personalized Recommendation Based on Data Analysis

This paper aims to explore the design and implementation of product demand forecasting and personalized recommendation system based on data analysis. Through the collection and analysis of a large number of user behavior data, an accurate demand prediction model is constructed, and based on this, a personalized recommendation algorithm is implemented. The paper first introduces the research background and purpose, and then elaborates the research method, process and results. The method based on data analysis can effectively predict the product demand and provide users with personalized recommendation services, so as to improve user satisfaction and market competitiveness of enterprises.

Mid-term electricity demand forecasting using improved multi-mode reconstruction and particle swarm-enhanced support vector regression

Balancing electricity supply and demand is crucial for China's energy transition and the stability of its electricity market. Accurate prediction of mid-term electricity demand plays a vital role in mitigating supply-demand imbalances, enabling policymakers and power plants to make informed decisions. This study proposes a novel hybrid model, EEMD-SE-PSO-SVR, to forecast mid-term electricity demand in China. The model integrates ensemble empirical mode decomposition (EEMD) with sample entropy (SE) for data preprocessing, particle swarm optimization (PSO) for parameter optimization, and support vector regression (SVR) for prediction. Our findings demonstrate that the EEMD-SE-PSO-SVR model outperforms traditional benchmark models, achieving a 54.49 % reduction in mean absolute percentage error (MAPE) compared to the SVR model. The model's performance is significantly enhanced by EEMD-SE, which effectively addresses the non-stationarity of electricity demand data. Moreover, the analysis highlights the strong influence of economic factors, followed by the seasonal factors and energy structure, underscoring their importance in accurately forecasting electricity demand. These findings contribute valuable insights for improving the accuracy of mid-term electricity demand forecasts and support the development of carbon-neutral and peak-carbon policies in China.