Estimation of posterior probabilities of consumer situational choices with neural network classifiers

This study shows how neural networks can be used to estimate the posterior probabilities in a consumer choice situation. We provide the theoretical basis for its use and illustrate the entire neural network modeling procedure with a situational choice data set from AT&T. Our findings supported the appropriateness of this application and clearly illustrate the nonlinear modeling capability of neural networks. The posterior probability estimates clearly add to the usefulness of the technique for marketing research.     

A score-driven model of short-term demand forecasting for retail distribution centers

Forecasting is one of the fundamental inputs to support planning decisions in retail chains. Frequently, forecasting systems in retail are based on Gaussian models, which may be highly unrealistic when considering daily retail data. In addition, the majority of these systems rely on point forecasts, limiting their practical use in retailing decisions, which often requires the full predictive density for decision making. The main contribution of this paper is the modeling of daily distribution centers (DCs) level aggregate demand forecasting using a recently proposed framework for non-Gaussian time series called score-driven models or Generalized Autoregressive Score (GAS) models. An experimental study was carried out using real data from a large retail chain in Brazil. A log-normal GAS model is compared to usual benchmarks, namely neural networks, linear regression, and exponential smoothing. The results show the GAS model is a competitive alternative to retail demand forecasting in daily frequency, with the advantage of producing a closed form predictive density by construction.     

A Model of the Determinants of Purchasing from Virtual Stores

Virtual stores are Internet-based innovations that influence the dynamics of consumer choice making. Utilizing the attitude-to-behavior theory, theory of reasoned action, technology acceptance model, functional theory of attitude, causal theory of action, and prior literature, we develop and empirically evaluate a model that describes consumer purchase decisions in a virtual store. We test the model using data collected from validated survey instruments for each of the constructs utilizing the structured equation modeling technique. The model helps in the design of virtual stores by describing how individuals who visit such stores can be induced to purchase from virtual stores. It describes that perceived usefulness, perceived behavioral control, and perceived peer influence impact attitude toward purchasing from a virtual store. Attitude toward purchasing from a virtual store, in turn, influences the actual purchasing from a virtual store. We discuss the implications for the design of virtual stores that lead to purchase decisions.     

Segmenting consumer decision-making styles (CDMS) toward marketing practice: A partial least squares (PLS) path modeling approach

The way consumers make decisions across online and offline channels according to their perceptions of retailers׳ marketing practices is not well understood in the current literature. A few empirical studies have examined consumer decision-making styles (CDMS) and marketing practice among online and shopping mall consumers. This research endeavors to understand CDMS as market segments on the perception of marketing practice across retail channels. A total of 315 online and paper–pencil-questionnaires were collected to conduct the statistical analysis for the measurement and structural model using the partial least squares (PLS) path modeling approach, a structural equation modeling (SEM) technique. Six consumers׳ characteristics–perfectionist, brand consciousness, price conscious, fashion conscious, recreational and impulsive shoppers–were assessed in respect of retail products, price, advertising and retailing across-channels. The study contributes to retailing management by enabling it to implement effective retail segmentation and cross-channel strategy according to CDMS. The study limitations and implications are discussed.     

一种新型的基于神经网络的无线信道模型

本文提出了一种使用神经网络来建模和仿真无线通信信道的方法。由于神经网络优异的学习特性,文中使用神经网络对无线信道的输入输出响应进行建模,讨论了建模的方法,给出了仿真的结果,分析了遇到的问题和解决方法,并进一步阐述了在实际的无线信道环境中使用神经网络建模的优越性。     

An empirical investigation and comparison of neural networks and regression for scanner data analysis

The objective of this study is to examine neural networks as an alternative to traditional statistical methods for the analysis of scanner data. The results of the study showed that neural networks can be an effective alternative to regression for modeling and predicting the effects of retailer activity on brand sales. The neural network models exhibited better performance in terms of both mean squared error and R² than the regression model.     

Enriching interactions: Incorporating outcome data into static discrete games

When modeling the behavior of firms, marketers and micro-economists routinely confront complex problems of strategic interaction. In competitive environments, firms make strategic decisions that not only depend on the features of the market, but also on their beliefs regarding the reactions of their rivals. Structurally modeling these interactions requires formulating and estimating a discrete game, a task which, until recently, was considered intractable. Fortunately, two-step estimation methods have cracked the problem, fueling a growing literature in both marketing and economics that tackles a host of issues from the optimal design of ATM networks to the choice of pricing strategy. However, most existing methods have focused on only the discrete choice of actions, ignoring a wealth of information contained in post-choice outcome data and severely limiting the scope for performing informative counterfactuals or identifying the deep structural parameters that drive strategic decisions. The goal of this paper is to provide a method for incorporating post-choice outcome data into static discrete games of incomplete information. In particular, our estimation approach adds a selection correction to the two-step games approach, allowing the researcher to use revenue data, for example, to recover the costs associated with alternative actions. Alternatively, a researcher might use R&D expenses to back out the returns to innovation.     

神经网络在客户流失模型中的应用研究

针对以往企业决策者难以掌握企业客户的流失倾向,运用神经网络的基本原理建立客户流失预测模型,使得决策者在商业实施过程中能够做到有的放矢,并且结合企业实例检测了该模型的可靠性。     

The perception of discount sales promotions – A utilitarian and hedonic perspective

This paper explores the effects of utilitarian and hedonic benefits perceptions of discount sales promotions on consumer purchase decisions (attitude and behavioral intention). A closed-ended survey, as well as a partial least squares structural equation modeling of consumers sampling who had used discount sales promotions in their recent purchases, was utilized in this study. The findings suggest that utilitarian benefits perceptions derived from price-quality perception and hedonic benefits perceptions derived from smart shopper self-perception of using discount sales promotions positively affect consumer purchase decisions. Both utilitarian and hedonic benefits perceptions of discount sales promotions can henceforth be used to produce the desired effects on purchase decisions. Hence, this paper sheds light on the effects of utilitarian and hedonic benefits perception of discount sales promotions on purchase decisions. Though the findings are limited to discount sales promotions alone and two types of purchase decisions outcomes (attitude and behavioral intentions), this paper identified multiple avenues to extend its findings. This encourages further exploration of such promotion topics, particularly on the various perceptions of discount sales promotions as well as other sales promotion techniques among marketing academics and practitioners.     

Positioning analysis with self-organizing maps: An exploratory study on luxury hotels

Positioning analysis examines the consumers' evaluative processes involving perceptions and preferences. Consumer behavior theory provides the hypotheses linking the perceptual and preferential positions of product brands or businesses to the expected brand choice decisions. For better visualization, various techniques of multivariate analysis assist in locating these positions in a space of low dimensionality. This data-processing step is necessary for bridging the gap between the fuzzy and redundant language of the consumer and the more abstract vocabulary of managers and service engineers. Hospitality services tend to be described with a rather large number of emotional and functional attributes. Condensing them into a compact spatial representation requires a significant data reduction effort. But the analyst faces still other challenges. Measuring perceptions and preferences turns out to be a segment-specific exercise. Strategic decisions on product positioning and market segmentation are interrelated. Contemporary decision-support models run the risk of becoming too complex to meet with managerial acceptance. Therefore, a simple method such as self-organizing maps (SOMs) may be attractive for exploratory purposes. SOMs are a variant of neural network models capable of detecting symptomatic brand positions. Managers may learn about typical combinations of product attributes, how they are assigned to brands, and how many and which consumers share similar views about brand profiles. A sample application of image and satisfaction measurement for luxury hotels demonstrates the technique. Strengths and weaknesses of this new neurocomputing approach to joint positioning/segmentation analysis are cautiously assessed. More experience from real-world applications is needed, and the industry is encouraged to experiment with SOMs on a larger scale.     

采用混合核支持向量机的DOA估计

人工神经网络（ANN）进行建模时通常需要准备大量的数据样本,同时网络结构一般都比较复杂;而采用支持向量机（SVM)进行建模时,不同核函数有不同的效果,各有利弊,且选取SVM模型参数的理论支撑尚不完整。为了解决这些问题,提出了一种基于混合核函数的支持向量机来改善来波到达角（DOA）的估计性能,并结合二进制粒子群算法（PSO）来对混合核函数进行参数寻优。该混合核函数由全局核函数和局部核函数构成,提高了SVM的泛化能力和学习能力。首先通过拟合多项式函数,验证了该混合核SVM的有效性。将该方法用于DOA估计建模,在不同信噪比和快拍数下,通过与径向基函数（RBF）神经网络、基于各单一核函数的SVM和MUSIC算法预测结果对比,混合核SVM均方差有所降低,提高了DOA估计的精度且有更好的稳定性。     

The impact of household level heterogeneity in reference price effects on optimal retailer pricing policies

The field of marketing has witnessed substantial improvement in modeling household level heterogeneity. However, relatively little has been written about how modeling household heterogeneity translates into better marketing decisions. In this paper, we study the impact of household level heterogeneity in reference price effects on a retailer's pricing policy. Reference prices are certain anchors or standards that households use to compare the observed purchase price of a product against. If the observed price is greater than the reference price it is perceived as a “loss” and if it is smaller than the reference price it is perceived as a “gain”. In order to study the impact of heterogeneity in reference price effects on retail pricing, we test a nested logit model under two alternative reference price (memory and stimulus based) and heterogeneity (finite mixture and hierarchical Bayes) specifications. In the empirical analysis, we find that households are quite heterogeneous in terms of their gain and loss effects. For some households a gain has higher impact than a corresponding loss, while the opposite is true for others. Using individual level estimates we then develop a normative pricing policy for a retailer maximizing category profit. Our results indicate that the optimal pricing policy derived from the heterogeneous case is qualitatively different, and more profitable, than the case when heterogeneity is ignored. We show that for an important marketing problem pertaining to a retailer, the optimal pricing decisions for various brands in a category are inextricably related to household heterogeneity in reference effects and brand preference.     

Towards a Strategic Model of Global Franchise Expansion

This paper develops the first global index of international franchise expansion that ranks countries according to their attractiveness to US-based franchise firms. A quantitative model combining insights from academic research and business practice generates a ranking of 143 potential expansion target countries according to their risk/opportunity profiles. The rankings suggest that countries with large markets and strong political and legal systems (large European countries, and Canada, Japan, and Australia) are the most attractive for US-based franchisors, while the small, unstable African countries are the least attractive. China (and the other BRIC countries—Brazil, Russia, India) though attractive from a market opportunity perspective, is nevertheless not highly ranked due to their significant risks and large cultural and geographic distances. This study reaffirms the importance of a strategic approach to international franchising decisions, underscores the importance of properly assessing the relative importance of key determinants in internationalization decisions, highlights the importance of a comprehensive and systematic assessment of the various risks in international franchising decisions, shows the usefulness of quantitative modeling in international franchising, and advocates the development of effective risk management methods in order to cope with rapid changes in the global marketplace.     

Cultural and Ethical Effects on Managerial Decisions: Examined in a Throughput Model

Financial and cost accounting information is processed by decision-makers guided by their particular need to support decisions. Recent technological advances impacting on information as well as organizations such as the European Community mandating financial reporting requirements for many countries is rapidly changing the landscape for decision making using accounting information. Hence, the importance of individuals' decision making is more important than it was previously. These decisions are also influenced by individuals' ethical beliefs. The Throughput Modeling approach to cultural and ethical concerns provides a way of dealing with accounting information processed through various pathways by decision-makers. This modeling approach captures different philosophical perspectives from which to understand what is involved in "thinking scientifically." In the Throughput Modeling approach, pathways highlight the importance of how different philosophical perspectives may be used by individuals in arriving at a decision. This paper highlights key concepts involved in rethinking the basis of moral decision making in terms of an underlying process, rather than focusing on the application of principles or the development of a virtuous character. Examples are provided from both English and Spanish settings to help emphasize the importance of modeling ethical decision making globally.     

Common Context for Decisions and their Implementations

Decisions are frequently sent to implementers without the context that guided them. It should not be a surprise, then, that results are not as expected. The lack of supplementary information and a common context produces wrongly implemented or lost decisions. This paper proposes a solution to this problem based on groupware technology. In particular, a combination of tools including shared workspaces, process modeling with workflow and a discussion tool, is proposed. A case is used to illustrate the problem and its solution.     

Optimal hedging with higher moments

This study proposes a utility‐based framework for the determination of optimal hedge ratios (OHRs) that can allow for the impact of higher moments on hedging decisions. We examine the entire hyperbolic absolute risk aversion family of utilities which include quadratic, logarithmic, power, and exponential utility functions. We find that for both moderate and large spot (commodity) exposures, the performance of out‐of‐sample hedges constructed allowing for nonzero higher moments is better than the performance of the simpler OLS hedge ratio. The picture is, however, not uniform throughout our seven spot commodities as there is one instance (cotton) for which the modeling of higher moments decreases welfare out‐of‐sample relative to the simpler OLS. We support our empirical findings by a theoretical analysis of optimal hedging decisions and we uncover a novel link between OHRs and the minimax hedge ratio, that is the ratio which minimizes the largest loss of the hedged position. © 2011 Wiley Periodicals, Inc. Jrl Fut Mark 32:909–944, 2012     

基于Petri Net和BPR的企业过程建模研究与应用

企业过程建模、分析和优化是企业过程重组BPR的基础,行之有效的建模技术是成功实施BPR的关键。为了对企业过程建模技术进行系统研究,在深入分析BPR特点的基础上,主要结合PetriNet技术,提出企业进行过程建模的阶梯层次方法,建立了BPPN模型,描述了Petri Net在企业过程中的形式化表示,并运用一个实例来验证模型的实用性,可以使管理者了解企业的生产细节和流程,更好地做出决策。     

Brand Loyalty and Price Promotion Strategies: An Empirical Analysis

Though brand loyalty has been studied extensively in the marketing literature, the relationship between brand loyalty and retail pricing strategies is not well understood. Designing promotion strategies involves two key decisions: the percentage reduction in price from the existing price point (depth), and the frequency with which a product is promoted. These decisions, in turn, are critically dependent upon how many consumers can be convinced to switch to a brand by temporarily reducing its price, and how many are instead brand loyal. Theoretical models of how the strength of brand loyalty influence optimal promotion strategies have been developed, but there are no rigorous tests of their hypotheses that take into account wholesale price variation. We test how brand loyalty impacts promotion strategies for two frequently purchased consumer packaged good categories. Our results confirm that retailers promote strong brands shallower and more frequently compared to brands with weak loyalty. Our results highlight the importance of carefully modeling wholesale prices when testing behavioral models on retail pricing.     

Linking Exploration with Exploitation through Relationship Memory*

Companies that are able to help clients make better decisions and enhance their business capabilities to flourish. However, there is often poor rapport between operating personnel who are in touch with end customers and executive personnel who make decisions. Through exploration and exploitation processes, new ideas and actions flow from the individual (i.e., operating personnel) to the organization (i.e., the rest of individuals and groups that form part of the organization). This paper examines the relative importance and significance of “the relationship memory” as a bridge between “exploration” and “exploitation” processes and its effects on the creation of “customer capital” through an empirical investigation of 139 small to medium‐sized enterprises in the Spanish optometry sector using structural equation modeling validated by factor analysis.     

Extentions of probabilistic perceptual maps with implications for competitive positioning and choice

Perceptual product positioning maps which are derived from probabilistic scaling models possess some distinct advantages over their deterministic counterparts. However, many probabilistic models still labor under a number of restrictive mathematical conditions. This paper describes an anisotropic space extension that alleviates some of these limitations by explicitly modeling the dimensional variances and covariances of each brand in a product positioning map. To clarify the decisions necessary when using probabilistic scaling models and to illustrate some of their attractive properties, two sets of convenience goods data are analyzed. The applications focus on the model's implications for the understanding of brand positioning and choice probabilities.