A support vector machine (SVM) approach to imbalanced datasets of customer responses: comparison with other customer response models

Customer response is a crucial aspect of service business. The ability to accurately predict which customer profiles are productive has proven invaluable in customer relationship management. An area that has received little attention in the literature on direct marketing is the class imbalance problem (the very low response rate). We propose a customer response predictive model approach combining recency, frequency, and monetary variables and support vector machine analysis. We have identified three sets of direct marketing data with a different degree of class imbalance (little, moderate, high) and used random undersampling method to reduce the degree of the imbalance problem. We report the empirical results in terms of gain values and prediction accuracy and the impact of random undersampling on customer response model performance. We also discuss these empirical results with the findings of previous studies and the implications for industry practice and future research.     

Attributes of the entrepreneurial type

In order to be entrepreneurs, people need to have both original ideas and the ability to make them work. This article focuses on three attributes which make this entrepreneurial orientation possible.     

A Cross-Cultural Comparison of Ethical Orientations and Willingness to Sacrifice Ethical Standards: China Versus Peru

Despite an increase in international business ethics research in recent years, the number of studies focused on Latin America and China has been deficient. As trade among Pacific Rim nations increases, an understanding of the ethical beliefs of the people in this region of the world will become increasingly important. In the current study 208 respondents from Peru and China are queried about their ethical ideologies, firm practices, and commitment to organizational performance. The empirical results reveal that Chinese workers are more relativistic and less idealistic than their Peruvian counterparts. One explanation for the disparity between these two groups is likely the variation in collectivism that can be traced to different levels of importance across ingroups and outgroups. In addition to a summary of the results, future research directions and managerial implications are discussed.     

The economics of a stock pollutant: Aldicarb on Long Island

This paper develops a dynamic model of groundwater contamination by the pesticide aldicarb on eastern Long Island. We estimate what the likely concentration would have been under static profit maximization and the marginal damage coefficient implied by the New York State health standard of 7 ppb. Based on our model, it appears that the concentration of aldicarb will not decline below 7 ppb until 1996 and that the shadow price of the current health standard is about $475,000 per year.     

An exploration of lesbian,gay, bisexual,and transgender pride festival sponsors

Pride festivals, events defined as a parade, march, rally, or festival organized for the lesbian, gay, bisexual, and transgender community, continue to rely on sponsorships for festival operations. The purpose of this exploratory study is to examine the current makeup of pride festival sponsors in the United States and Canada. A content analysis of 780 of 30 U.S. and Canadian pride festival sponsors found that all 30 pride festivals utilized sponsors. Findings indicate that a majority of the sponsors that came from states outside of the pride festival state were private corporate sponsors, and came from industries such as services, health, media, and beverages. Theoretical and managerial implications are discussed.     

Social indicators and social accounts

This paper defines “social indicators” as statistics which have two denning characteristics. They are, first, measures of direct normative interest; that is, what the economist would call measures of “welfare” and “illfare”. Most existing government statistics are not of this type, because a large proportion of existing statistics are measures of government or other institutional activity, produced as a by-product of accounting or administrative routine. The second defining characteristic of a social indicator is that it should fit into a systematic scheme of classification or aggregation which would make possible a balanced assessment of socio-economic progress or retrogression in some broad area, as well as disaggregated and detailed study of particular problems.

The work in government on social indicators was designed in part to meet the needs of Toward a Social Report, a preliminary study of the condition of American society issued by the U.S. Department of Health, Education, and Welfare (Superintendent of Documents, Wash. D.C. 20402; 1969). Social indicators can also fit, with other statistics, into a set of “policy accounts” or scheme of social accounting, which would relate social expenditures to the social indicator they were designed to affect. This would encourage broadened cost-benefit analysis and rational public decision-making.     

A technique using analytical hierarchy process in multiobjective planning models

Social and economic planning usually includes the need to consider and reconcile multiple, conflicting objectives. A number of multiple objective planning models have been proposed, using goal programming as well as other multiple objective techniques. Goal programming provides a useful tool which quickly provides a solution matching the set of ranked objective targets provided, but does not provide much exploratory analysis of planning problems. Other techniques have been used which provide generation of new alternatives for consideration, but these other techniques often are limited to original model extreme points. This paper presents an alternative means of generating planning alternatives, utilizing Analytic Hierarchy Process as a means of obtaining a good starting point, and imposition of objective attainment levels through constraints. The proposed process provides a means of decision maker learning, easily allows adjustment of objective attainment levels and is not limited to the original model extreme points.