基于Stacking模型融合的店铺销量预测

为了实现企业产品销量预估,提高生产供应的准确性与效率,提出了基于Stacking模型的融合算法进行销量预测。算法设计了两层堆叠的模型结构,初级学习器采用随机森林、支持向量回归、差分整合移动平均自回归、轻量级梯度提升机器和门控循环单元5种单模型,将分类与回归树作为次级学习器构成Stacking融合模型,并对数据进行了预测。预测结果显示,使用Stacking模型融合后得到了较好的预测结果,比单模型中效果最好的模型的均方根误差更小,平均绝对误差更小,决定系数值更大,表明Stacking融合模型的预测准确率更高。所设计模型可用于对企业店铺的产品销量进行预测,帮助企业更好地安排生产、营销活动,为减少库存、缩短生产销售周期提供数据支持,对企业生产决策有一定的参考价值。     

基于回归-ELM神经网络模型的滑坡变形及失稳预测模型

为准确预测滑坡的变形趋势,有效预防滑坡灾害的发生,提出了基于变形预测和检验的趋势判断模型。首先,利用回归分析,拟合得到滑坡的变形曲线,再利用组合权值,实现拟合结果的组合,得到滑坡变形的初步预测结果;其次,利用极限学习机(ELM神经网络)对初步预测结果进行误差修正,将修正结果与初步预测结果进行叠加,得到滑坡变形的综合预测值;最后,利用秩相关系数检验与Mann-Kendall检验,对滑坡变形趋势进行判断,以验证预测结果的准确性。经过实例检验得出,预测模型的预测效果较好,其组合预测及误差修正均能不同程度地提高预测精度及稳定性,且两检验模型的结果均与预测结果相符,相互验证了其可靠性。因此,预测模型能对滑坡变形趋势进行综合判断,为滑坡的变形研究提供了一种新的思路。     

Robust Testing of the Utility‐based High‐technology Product Sales Forecasting Models Proposed by Decker and Gnibba‐Yukawa (2010)

Decker and Gnibba‐Yukawa (2010) propose an elegant utility‐based model for forecasting the sales of high‐technology products and suggest that the model yields forecasts that are highly accurate. However, this finding is based on forecasts for a total of only six holdout observations shared across three products. This number of observations is insufficient for reliable inferences to be drawn about the accuracy of a method and the use of such a small data set runs counter to an accepted principle of forecast evaluation. The authors’ proposed model was tested on more extensive data and sensitivity analysis applied to the results. No evidence was found that the utility‐based model could outperform a relatively simple extrapolative model despite the much greater effort involved in applying the proposed model. In addition, the utility‐based model is only applicable for forecasting sales during a narrow interval in a product's life cycle and requires several periods of historic sales data before it can be implemented. It also depends heavily on the accurate estimates of parameters that are determined outside the model (and which may depend on difficult judgments by managers) and assumes that consumers or households will only purchase the product once between the launch date and the forecast horizon. In light of this, it is argued that the utility‐based model is likely to have limited usefulness as a sales forecasting tool.     

An Evaluation of Hedonic Price Indexes for Thirty-four Large SMSAs

The purpose of this paper is twofold. First, a set of price indexes are developed for thirty-four metropolitan areas for the years 1975 and 1978. Indexes are also broken down by central city and suburb for twenty two of the metropolitan areas. These indexes measure variation in the price per unit of owner-occupied housing stock and the price per unit of housing services of rental housing. Confidence intervals are constructed for the price indexes. It is found that the indexes are reasonably precise measures in that the typical ratio of the confidence interval to the predicted median rent or value is about 13%. Also, analysis of the confidence intervals indicates substantial and statistically significant variation in the price of housing stock and services among the metropolitan areas studied. One of the major benefits of this set of indexes is that it can be used to address one of the most important questions in real estate — Why do housing prices and rents vary among metropolitan areas and over time? Research projects are currently underway that use the data to address these two questions.     

Mortgage Default Risks and the 1971 San Fernando Earthquake

The paper is based on a study of mortgage default risks associated with natural disasters. These risks are faced by holders of mortgages when forced by default to acquire damaged properties. A sample of residential mortgage properties damaged in the 1971 San Fernando, California earthquake is studied. Some of the mortgagors defaulted while others did not. The paper identifies and analyzes those variables associated with default using discriminant and probit regression analysis. The study concentrates on earthquake exposures in California, but has implications for all major disasters.     

Analysis of the Characteristics of Individual Investors in Real Estate Securities and Income-Producing Property

Analysis of the responses to a nationwide survey of investors demonstrates that individuals who invest in real estate differ in a predictable way from those who invest in assets other than real estate. Two types of real estate investment vehicles are studied: income-producing (rental) property and real estate securities. A multiple group discriminant analysis model is presented which successfully classifies prospective investors into four groups (owners of income property only, real estate securities only, both, neither) with a predictive accuracy more than double the chance probability of correct classification. The results provide insights useful in policy analysis and the design and marketing of real estate investments.     

Property Taxes and Residential Rents

Property taxes are a fundamental source of revenue for local governments, constituting 73% of local government tax revenue in the United States. In this article, we empirically investigate the impact of residential property taxes on residential rents. Using data from the American Housing Survey and the National League of Cities, we estimate numerous specifications of a hedonic rent equation with comprehensive unit-level, neighborhood-level and city-level controls. We find that a one standard deviation increase in the property tax rate raises residential rents by roughly $400 annually.     

Statistical Approaches to Forecasting Domestic Energy Consumption and Assessing Determinants: The Case of Nordic Countries

The residential sector accounts for large share of total annual energy use in the Nordic countries due to the extremely cold climates and high household heating demand. Most domestic energy consumption in the Nordic countries is for space heating and providing hot water. The purpose of our study was to forecast the annual energy consumption of the Nordic residential sectors by 2020 as a function of socio-economic and environmental factors, and to offer a framework for the predictors in each country.

Our research models the domestic energy use in Nordic countries based on social, economic and environmental factors. Applying the multiple linear regression (MLR), multivariate adaptive regression splines (MARS), and the artificial neural network (ANN) analysis methodologies, three models have been generated for each country in the Nordic region. Using these models, we forecasted the Nordic countries domestic energy use by 2020 and assessed the causal links between energy consumption and the investigated predictors. The results showed that the ANN models have a superior capability of forecasting the domestic energy use and specifying the importance of predictors compared to the regression models. The models revealed that changes in population, unemployment rate, work force, urban population, and the amount of CO₂ emissions from the residential sectors can cause significant variations in Nordic domestic sector energy use.     

Uses and Abuses of Residential Building Permits in Forecasting Private Housing Starts

This paper examines the commonly held view that the Census Bureau's monthly series for residential building permits issued is a leading indicator of the series for private housing starts. A close study of the residential building permits and housing starts series shows that on a quarterly basis the series are coincident, with cyclical peaks and troughts in the seasonally adjusted data occurring in the same period. On the monthly basis for the period 1960 through 1976, the peaks and troughs in the building permits series have usually followed those in the housing starts series. Also, evidence presented in the paper shows that monthly changes in the residential building permits series are not highly correlated with changes in the housing starts series one to five months in the future. Finally, the analysis suggests that changes in the backlog of unused residential building permits are not indicative of future changes in the rate of housing starts. The major implication of these results is that permits data offer little or no information about future month-to-month changes in the rate of housing starts. However, because the building permits series is less irregular than the housing starts series and the cyclical movements of the series are basically the same, permits data can be used to identify irregular monthly movements in the starts series.     

余能原理基面力元法在重力坝应力分析中的应用

介绍了二维线弹性问题的余能原理基面力元法。该方法以基线力表征物体的受力状态,以其对偶量位移梯度表征物体的变形状态。通过重力坝算例与平面4节点等参元(Q4模型)和平面4节点减缩积分单元(Q4R模型)的对比分析,探讨了基面力元法应用于重力坝应力分析的可行性。数值分析结果表明,基于余能原理的基面力元法具有较高的计算精度,可以应用于重力坝应力分析问题,有较广阔的应用前景。     

Do it separately or simultaneously—An empirical analysis of a conjoint implementation of TQM and TPM on plant performance

Over the last decades, Total Quality Management (TQM) and Total Productive Maintenance (TPM) have become key concepts for improving production. The majority of modern manufacturing companies has implemented at least one of these improvement programs or even both of them. The common ground of TQM and TPM is the focus on human resources. In this paper, the impact of TQM and TPM on plant performance and especially the supporting role of employee involvement practices is investigated empirically using multiple regression analysis and structural equation modeling. The data used for conducting the analyses is taken from the international research project High Performance Manufacturing that contains the data of 238 plants. The results indicate that TQM and TPM, supported by HR practices, have a significant potential to improve plant performance. However, a simultaneous implementation of both concepts does not necessarily lead to superior performance. As potential reason for this, human resources are regarded as limiting factor both improvement programs draw on. Accordingly, this scarce resource is identified as crucial element with respect to performance when implementing TQM and TPM simultaneously.     

Examining salesperson versus sales manager evaluation of customer opportunities: A psychological momentum perspective on optimism,confidence, and overconfidence

Making correct decisions regarding the merit of sales opportunities, as well as deciding which opportunities to pursue and which to abandon, is one of the most critical tasks sales managers and salespeople grapple with. These frontline assessments typically become sales forecasts that are used to allocate resources within the firm. Assessments of future sales, however, are often inaccurate and little is known about how salespeople and sales managers make decisions about sales opportunities, or if they differ in their ability to (correctly) assess the merits of such opportunities. To address this research gap, we explore several decision-making biases – namely optimism, confidence, and overconfidence – through the lenses of both Psychological Momentum (PM) and Strategic Reference Point (SRP) theory. Analysis of rich CRM data from a Fortune 500 medical products company, across three studies, demonstrates notable differences between salespeople and sales managers concerning decision-making biases that influence their evaluations of sales opportunities.     

The accuracy of project cost and duration estimates in industrial R&D

Abstract . This study of the accuracy of cost and duration estimates, both initial and intermediate, for industrial R&D projects, is based mainly on analysis of the records of 475 projects in four varying research organizations. Mean ratios of actual to estimated cost ranging from 0.97 to 1.51 are obtained, and mean ratios of actual to estimated duration from 1.39 to 3.04. The results are generally similar to those of other British and American studies, with which they are compared.
There is no evidence that the information gained as projects progress enables their future cost and duration to be estimated more accurately; at best, the accuracy of such estimates remains constant. No effect of project size on estimate accuracy is found; effects of project length on accuracy, and of time (i.e. increasing experience) on accuracy are found only in one organization each. The pattern of expenditure over time is examined, and found on average to be not far from linear, though with wide individual project variations.
When the individual project ratios are reduced by constant factors representing optimistic or pessimistic bias (derived from the mean ratios, and assumed to be characteristic of the firm), and then subjected to a log transformation to make their distribution more symmetrical, the remaining variation, which is a measure of the inaccuracy of estimation for the individual projects, is closely similar in the four organizations, and not greatly different in the organizations covered by other studies when the data are similarly treated.     

Did federal regulation discourage facilities-based entry into US local telecommunications markets?

In the United States from 2001 to 2006, federal regulations allowed entrants to lease from incumbents at relatively low cost all of the network infrastructure necessary to provide local phone service. These platform entrants could then provide phone service without installing any of their own equipment. Advocates of this policy claimed that it was needed to provide an economically feasible means by which entrants could serve residential customers. Critics contended that the policy substantially deterred loop entry whereby entrants installed their own switching equipment. An analysis of panel data for each state over this period indicates that the policy's critics may have been correct. The cross-price elasticity of loop entry with respect to platform price was roughly 1.0. A back of the envelope calculation suggests that loop entry may have decreased by roughly 20% due to platform entry price reductions.     

基于多元线性回归的装饰工程快速估价模型

针对目前装饰工程陕速估价方法的不足，以及装饰工程造价组成的特点，建立一种基于多元线性回归的快速估价新模型。以住宅类建筑为例，建立其多元线性回归模型，结合SPSS软件对样本分析，结果表明，其线性回归模型显著，同时，内墙、外墙、楼地面和窗的回归系数也非常显著。通过计算回归值与实际值的误差，发现误差比较理想，该模型具有很好的实用价值。     

The determinants of entry and exit reconsidered

In this paper we empirically examine the hypothesis that there is a symmetrical relationship between entry and exit barriers. The analysis, using Canadian cross-section data, proceeds in two stages. In the first stage we estimate entry and exit equations which ignore the possibility that displacement occurs (that entry causes exit). These results indicate that there is symmetry in the determinants of entry and exit, but that the exit equation is misspecified. In the second stage, we estimate a model in which entrants are allowed to displace incumbents (cause exit). These results indicate much less symmetry primarily because in the absence of entry barriers, entry occurs and incumbents are displaced. We conclude that symmetry exists, but in the ex-ante sense that barriers to exit are barriers to entry.     

替代原理在房地产估价中运用

在简要介绍替代原理及其作用的基础上,分析了替代原理在房地产估价各种办法中运用情况。指出替代原理是指导房地产估价的普遍性原理,是房地产估价的重要原则,在房地产估价中具有广泛的应用。     

An Empirical Analysis of Hedonic Regression and Grid-Adjustment Techniques in Real Estate Appraisal

Multiple regression analysis has become increasingly popular when appraising residential properties for tax purposes. Alternatively, most fee appraisers and real estate brokers use the traditional sales comparison approach. This study combines the two techniques and uses multiple regression to generate the adjustment coefficients used in the grid adjustment method. The study compares the combined grid-regression method with ordinary regression and defines the market conditions under which each method is likely to be more effective. The grid-regression method is found to be more accurate for relatively homogeneous housing markets, and the multiplicative percentage adjustment method (MPAM) the preferred approach.     

The Debiasing of Forecasts in Research and Development

Abstract . A possible way to improve the accuracy of forecasts is to estimate (and subsequently remove) any systematic bias by reference to the bias found in similar forecasts in the past. This paper discusses the applicability to industrial R & D of a method proposed by Summers for debiasing estimates of the costs of military projects by means of a regression equation derived from past experience, and gives the results of applying a similar method to estimates of costs, sales, product prices, and return factors for ten industrial development projects.     

制药企业环境绩效的统计测度与影响因素分析

本文首先运用 DEA-Malmquist 方法分析“十二五”期间制药业企业的环境绩效,再引入Bootstrap-DEA 方法进行纠偏并给出置信区间,并选用分位数回归对制药业企业环境绩效进行影响因素分析。得出结论：“十二五”期间制药业企业环境绩效的改善,主要是技术进步的作用。分位数回归显示企业的经济规模、企业技术成熟度、企业的环保意识及公共压力都会影响企业环境绩效的改善,且分位数回归全面展示了各因素的作用属性,所得到的结论更符合现实。