Classification-based model selection in retail demand forecasting

Retailers supply a wide range of stock keeping units (SKUs), which may differ for example in terms of demand quantity, demand frequency, demand regularity, and demand variation. Given this diversity in demand patterns, it is unlikely that any single model for demand forecasting can yield the highest forecasting accuracy across all SKUs. To save costs through improved forecasting, there is thus a need to match any given demand pattern to its most appropriate prediction model. To this end, we propose an automated model selection framework for retail demand forecasting. Specifically, we consider model selection as a classification problem, where classes correspond to the different models available for forecasting. We first build labeled training data based on the models’ performances in previous demand periods with similar demand characteristics. For future data, we then automatically select the most promising model via classification based on the labeled training data. The performance is measured by economic profitability, taking into account asymmetric shortage and inventory costs. In an exploratory case study using data from an e-grocery retailer, we compare our approach to established benchmarks. We find promising results, but also that no single approach clearly outperforms its competitors, underlying the need for case-specific solutions.     

Limited information Bayesian analysis of a simultaneous equation with an autocorrelated error term and its application to the U.S. gasoline market

Using Markov Chain Monte Carlo algorithms within the limited information Bayesian framework, we estimate the parameters of the structural equation of interest and test weak exogeneity in a simultaneous equation model with white noise as well as autocorrelated error terms. A numerical example and an estimation of the supply and demand equations of the U.S. gasoline market show that if we ignore autocorrelation we obtain unreasonable posterior distributions of the parameters of interest. Also we find that the hypothesis of the asymmetric effect of the changes in oil price on the changes in gasoline price is rejected. Oil inventory has a significant negative effect on the gasoline price.     

The economic cost of fuel subsidies in Ghana

Apart from fuel subsidies failing to provide the greatest benefits to the poor, they create overconsumption, market inefficiencies and negative environmental and social externalities. This study extended the comprehensive Harberger formula for deadweight loss (DWL) to account for the cross-price substitution effects between the gasoline and diesel markets, estimated the own-price and cross-price elasticities for gasoline and diesel in Africa, and predicted the DWL or economic cost created by gasoline and diesel subsidies in Ghana. Using the simple Harberger formula yields DWL estimates ranging from GHS 2.03 to 8.58 million per year from 2009 to 2014, but accounting for cross-price effects reduces the DWL estimates, from GHS 1.53 to 7.55 million per year, over the same period. However, because fuel demand is highly price inelastic, the DWL created for every GHS 1 million spent on gasoline and diesel subsidies only represents between 0.5% and 2% of subsidy expenditures. These findings suggest that fuel subsidy reforms would be better motivated by other social problems associated with fuel subsidies, such as the ineffective targeting of such subsidies to poorer consumers and the negative externalities of overconsumption, rather than the economic inefficiencies that fuel subsidies generate.     

Google data in bridge equation models for German GDP

Interest in the use of “big data” when it comes to forecasting macroeconomic time series such as private consumption or unemployment has increased; however, applications to the forecasting of GDP remain rather rare. This paper incorporates Google search data into a bridge equation model, a version of which usually belongs to the suite of forecasting models at central banks. We show how such big data information can be integrated, with an emphasis on the appeal of the underlying model in this respect. As the decision as to which Google search terms should be added to which equation is crucial —- both for the forecasting performance itself and for the economic consistency of the implied relationships —- we compare different (ad-hoc, factor and shrinkage) approaches in terms of their pseudo real time out-of-sample forecast performances for GDP, various GDP components and monthly activity indicators. We find that sizeable gains can indeed be obtained by using Google search data, where the best-performing Google variable selection approach varies according to the target variable. Thus, assigning the selection methods flexibly to the targets leads to the most robust outcomes overall in all layers of the system.     

Modeling voter choice to predict the final outcome of two-stage elections

Most election forecasting research to date has been conducted in the context of single-round elections. However, more than 40 countries in the world employ a two-stage process, where actual voting data are available between the first and the second rounds to help politicians understand their position in relation to each other and to voter preferences and to help them predict the final outcome of the election. In this study we take advantage of the theoretical foundation on voter behavior from the political science literature and the recent methodological advances in choice modeling to develop a Nested Logit Factor Model of voter choice which we use to predict the final outcome of two stage elections and gain insights about the underlying political landscape. We apply the proposed model to data from the first stage and predict the final outcome of two stage elections based on the inferences made from the first stage results. We demonstrate how our proposed model can help politicians understand their competitive position immediately after the first round of actual voting and test its predictive accuracy in the run-off election across 11 different state governorship elections.     

Autoverkopen en autopark in Nederland, 1950–1970

Summary This is a study of the demand for the ownership of new or firsthand cars, i.e. the demand of ‘first owners’ who habitually buy new cars which are then traded in long before they are obsolescent. Upon combining the evidence of various surveys with time-series for 1950-64 it is found that this ownership has an income elasticity of 2 and a price elasticity of – 1.25 in respect of the constant-quality index of new car prices earlier provided. On the average new cars are traded in after three years so that roughly one third of the existing stock is replaced every year. These values together determine an equation for purchases of new cars which accurately predicts new registrations in 1965 and 1966. Projections for the years 1967-70 are provided. In the model employed the total number of cars (including used cars) is determined by the active demand exerted by ‘first owners’ on one hand and by the hitherto fairly constant scrappage rates on the other. Projections of the overall ownership rate can therefore be derived from the forecasts of new registrations. The ownership rate will approach 80% around 1970, and it is likely that at that stage the current scrappage rates will cease to apply in view of the greater predilection for comparatively younger cars.     

Long-term forecasting of fuel demand at theater entry points

Managing the distribution of fuel in theater requires Army fuel planners to forecast demand at the strategic level to ensure that fuel will be in the right place, at the right time, and in the amounts needed. This work presents a simulation approach to forecasting that accounts for the structure of the supply chain network when aggregating the demand of war fighters across the theater over the forecasting horizon. The resulting empirical distribution of demand at the theater entry point enables planners to identify forecast characteristics that impact their planning process, including the amplitudes and temporal positions of peaks in demand, and the estimated lead time to the point of use. Experimentation indicates that the forecasts are sensitive to the pattern of war fighter demand, the precise structure of the in-theater supply chain network, and the constraints and uncertainty present in the network, all of which are critical planning considerations.     

An empirical analysis of effects of gasoline price change on transportation behavior in Japan,with consideration of regional differences

This empirically analyzes the effect of gasoline price change on various aspects of surface transport behavior comprehensively, with consideration of regional differences as to whether or not the region includes a major metropolitan agglomeration area (hereafter, “megalopolis”), by means of carefully arranging yearly time-series data from fiscal 1987 to fiscal 2008 in Japan. The aim is to look at common as well as different effects among regions and to draw policy implications relating to fuel price, as well as to fuel tax, towards developing a low-carbon transport system, which are applicable to other countries including emerging countries where a demographic disparity between megalopolis areas and rural areas might be increasing. The methodology includes the multiple regression models complemented by the 1st order auto-regressive models of error terms, where the elasticities of gasoline price and those of income regarding usage of personal automobiles, of public transport and of passenger railways with or without periodic tickets, and fuel consumption on cars, as well as ownership of various types of cars, are estimated for individual gross-regions. In both gross-regions, higher gasoline prices are commonly shown to be related to lower ownership of larger private passenger cars, higher ownership of light cars, lower ownership of the total private passenger cars, lower passenger-km per capita in cars, lower fuel consumption on cars and higher transit ridership. The result suggests general automobile-fuel-price policy implications, the extent of which is differentiated by spatial characteristics, and estimates the adverse effect of the possible abolition of the provisional gasoline tax rate on the emissions in both gross-regions. The application of result is also tried in exploring effect of other policy measures such as expressway tolls.     

A fast and scalable ensemble of global models with long memory and data partitioning for the M5 forecasting competition

This work presents key insights on the model development strategies used in our cross-learning-based retail demand forecast framework. The proposed framework outperforms state-of-the-art univariate models in the time series forecasting literature. It has achieved 17th position in the accuracy track of the M5 forecasting competition, which is among the top 1% of solutions.     

Measuring residential energy efficiency improvements with DEA

This paper measures energy efficiency improvements of US single-family homes between 1997 and 2001 using a two-stage procedure. In the first stage, an indicator of energy efficiency is derived by means of Data Envelopment Analysis (DEA), and the analogy between the DEA estimator and traditional measures of energy efficiency is demonstrated. The second stage employs a bootstrapped truncated regression technique to decompose the variation in the obtained efficiency estimates into a climatic component and factors attributed to efficiency improvements. Results indicate a small but significant improvement of energy efficiency over the studied time interval, mainly accounted for by fuel oil and natural gas users.     

乙醇汽油对发动机性能影响的实验研究

随着国际石油供应越来越紧张,以及汽车尾气排放污染越来越严重,乙醇汽油已经成为今后可再生资源的发展重点。目前,我国很多地区也已经开始试用乙醇汽油,但是乙醇汽油毕竟是新生事物,它对汽车发动机的性能到底存在哪些影响,我们至今还没有一个确切的结论。本文通过一系列实验,对乙醇汽油对发动机性能的影响进行了探讨。     

Regional forecasting of the demand for fossil fuels by electric utilities in the United States

The possibility of interfuel substitution in the generation of electrical energy exists. Given this fact, a demand model for various fossil fuels by electric utilities in the United States is developed. Using regional data, the results suggest that the responsiveness of the demand for coal, residual fuel oil, distillate fuel oil, and natural gas by electric utilities to relative price changes is significant. In a forecasting setting, the demand model performs remarkably well when actual and forecast values for 1979 are compared.     

MTBE行业供需现状及市场前景

近年来在汽油高标号需求大幅增长的推动下,国内MTBE产量随着产能大幅增长,但仍未能满足国内MTBE的需求,MTBE市场整体呈现"供不应求"局面,依靠部分进口资源填补国内的空缺。后期随着国内汽油质量的不断升级及调和技术的不断成熟,MTBE用于调和的量将会不断增多。另外MTBE用于化工各行业存在生产高技术含量、附加值高、需求稳定,今后对MTBE的消费量将会有较大的增加,具有较好的市场前景。     

The impact of forecasting on operational performance: Mediation effects through flexibility enablers

Literature has devoted increasing attention to the problem of supply and demand management in uncertain contexts. Only limited contributions, however, can be found regarding the interaction between forecasting and flexibility enablers to manage demand as well as regarding the flexibility enablers’ effect on company performance. We will discuss the impacts of flexibility and forecasting on dynamic interactions. The aim of this work is to study the mediation effect of forecasting through flexibility enablers on company performance, i.e., customer satisfaction and cost efficiency. Our results provide evidence that the relationship between forecasting and customer satisfaction is mainly due to process flow management, while the relationship with cost efficiency is mainly due to layout.     

Evaluating spatial service and layout efficiency of municipal Wi-Fi facilities for SmartCity planning: A case study of Wuhan city,China

This paper explores an approach to evaluating spatial service and layout efficiency of municipal Wi-Fi facilities for cyber infrastructure planning in the construction of a smart city. Based on the equilibrium principle that Wi-Fi services supply should meet the spatial demand for recreation, urgency, and security, two evaluation models were built. In the first model, the spatial service efficiency of Wi-Fi facilities is measured by a mean ratio of three types of demand spaces covered by Wi-Fi service. In the other model, spatial layout efficiency of each Wi-Fi facility is calculated by a demand load indicator - the sum of the three types of demand space's ratios of served quantity to the average quantity it should carry, and then assessed by relationship with its adjacent facilities. Corresponding statistical methods and analysis processes were also designed using the function modules found in ArcGIS software. The approach was applied to Wuhan city in 2016. After overlaying influence areas of 688 Wi-Fi facilities with demand spaces of 137 access points for recreation, 1200.99 km roads for urgency, 121.54 km² regions for security, the results show that there exists a spatial disequilibrium between supply and demand of Wi-Fi service in Wuhan due to lack of comprehensive planning. Specifically, there are 67 points, 700.76 km roads, and 42.59 km² regions of demand spaces covered by Wi-Fi service, with service efficiencies of 48.91%, 58.35%, and 35.04% respectively and an overall spatial service efficiency of 47.43%. Among all the Wi-Fi facilities, there are 481 carrying saturated demand, 3 straying from the centre of demand space, and 204 redundant. The overall spatial layout efficiency of Wi-Fi facilities is 43.01%. The conclusion is that the dense and massive redundant Wi-Fi facilities should be optimized, especially in the centre areas along the confluence of Yangtze and Han rivers and new Wi-Fi facilities supplemented in other ill-equipped areas. Since many metropolitan cities experience the same problem, this novel approach will find wide application in the future and offer improved evaluation strategies for researchers and policymakers.     

Determining optimal fuel delivery strategies under uncertainty

During the preparation before a hurricane makes landfall, affected individuals may be asked to evacuate. Large and small-scale evacuations can cause rapid increases in the demand for gasoline fuel. However, during a hurricane vessels carrying gas may be delayed and/or rerouted, adding to the difficulty of providing the necessary gas in affected areas. In this work, we determine alternate delivery locations and times for vessels carrying fuel that are scheduled to arrive and deliver fuel at ports impacted by an approaching hurricane. Motivated by Hurricane Irma in Florida, we develop a multi-period stochastic scheduling model that incorporates hurricane (weather) advisories, fuel delivery schedules, port storage capacities, and port docking capacities. Our model determines the best schedule based on two objectives: (1) minimize the total unmet demand at each port, and (2) minimize inequities in unmet demands among the ports. We also present a case study and a numerical experiment based on fuel delivery data from ports in Florida. Among our key findings is that port availability is the driving factor in determining feasible schedules for vessel gas deliveries. We also present a scheduling heuristic that dynamically adapts to weather advisories so as to minimize the impact of unmet demand in the affected areas.     

Improving demand forecasting in the air cargo handling industry: a case study

Air transportation plays a crucial role in the agile and dynamic environment of contemporary supply chains. This industry is characterised by high air cargo demand uncertainty, making forecasting extremely challenging. An in-depth case study has been undertaken in order to explore and untangle the factors influencing demand forecasting and consequently to improve the operational performance of an air cargo handling company. It has been identified that in practice, the demand forecasting process does not provide the necessary level of accuracy, to effectively cope with the high demand uncertainty. This has a negative impact on a whole range of air cargo operations, but especially on the management of the workforce, which is the most expensive resource in the air cargo handling industry. Besides forecast inaccuracy, a range of additional hidden factors that affect operations management have been identified. A number of recommendations have been made to improve demand forecasting and workforce management.     

A modified Teunter-Syntetos-Babai method for intermittent demand forecasting

Intermittent demand refers to the specific demand pattern with frequent periods of zero demand. It occurs in a variety of industries including industrial equipment, automotive and specialty chemicals. In some industries or some sectors of industry, even majority of products are in intermittent demand pattern. Due to the usually small and highly variable demand sizes, accurate forecasting of intermittent demand has always been challenging.However, accurate forecasting of intermittent demand is critical to the effective inventory management. In this study we present a band new method - modified TSB method for the forecasting of intermittent demand. The proposed method is based on TSB method, and adopts similar strategy, which has been used in mSBA method to update demand interval and demand occurrence probability when current demand is zero. To evaluate the proposed method, 16289 daily demand records from the M5 data set that are identified as intermittent demands according to two criteria, and an empirical data set consisting three years’ monthly demand history of 1718 medicine products are used. The proposed mTSB method achieves the best results on MASE and RMASE among all comparison methods on the M5 data set. On the empirical data set, the study shows that mTSB attains an ME of 0.07, which is the best among six comparison methods. Additionally, on the MSE measurement, mTSB shows a similar result as SES, both of which outperform other methods.     

信息分享对原始设备制造商与其合同制造商竞争的影响

研究了原始设备制造商的预测信息分享对一个原始设备制造商和一个与其同时有合作和竞争的合同制造商组成的供应链系统的影响,建立制造商间信息分享的模型,该模型包括一个原始设备制造商和一个合同制造商。研究发现,原始设备制造商关于市场潜在需求预测信息的分享对其预期利润是不利的,同时需求信息预测的精度对原始设备制造商信息分享的决策也有影响,原始设备制造商没有动机与其供应链成员进行信息分享,但信息分享使得供应链整体利润增加。最后,建立一个信息分享补偿机制分享供应链利润的增加量,以期通过信息分享补偿机制促使原始设备制造商有动机进行信息分享,从而实现其与合同制造商的“双赢”。