Juan Manuel Ramon-Jeronimo

Marketing Segmentation Through Machine Learning Models

Customer relationship management (CRM) aims to build relations with the most profitable clients by performing customer segmentation and designing appropriate marketing tools. In addition, customer profitability accounting (CPA) recommends evaluating the CRM program through the combination of partial measures in a global cost—benefit function. Several statistical techniques have been applied for market segmentations although the existence of large data sets reduces their effectiveness. As an alternative, decision trees are machine learning models that do not consider a priori hypotheses, achieve a high performance, and generate logical rules clearly understood by managers. In this article, a three-stage methodology is proposed that combines marketing feature selection, customer segmentation through univariate and oblique decision trees, and a new CPA function based on marketing, data warehousing, and opportunity costs linked to the analysis of different scenarios. This proposal is applied to a large insurance marketing data set for alternative cost and price conditions showing the superiority of univariate decision trees over statistical techniques.

Modelling Credit Risk with Scarce Default Data: On the Suitability of Cooperative Bootstrapped Strategies for Small Low-Default Portfolios

Credit risk models are commonly based on large internal data sets to produce reliable estimates of the probability of default (PD) that should be validated with time. However, in the real world, a substantial portion of the exposures is included in low-default portfolios (LDPs) in which the number of defaulted loans is usually much lower than the number of non-default observations. Modelling of these imbalanced data sets is particularly problematic with small portfolios in which the absence of information increases the specification error. Sovereigns, banks, or specialised retail exposures are recent examples of post-crisis portfolios with insufficient data for PD estimates, which require specific tools for risk quantification and validation. This paper explores the suitability of cooperative strategies for managing such scarce LDPs. In addition to the use of statistical and machine-learning classifiers, this paper explores the suitability of cooperative models and bootstrapping strategies for default prediction and multi-grade PD setting using two real-world credit consumer data sets. The performance is assessed in terms of out-of-sample and out-of-time discriminatory power, PD calibration, and stability. The results indicate that combinational approaches based on correlation-adjusted strategies are promising techniques for managing sparse LDPs and providing accurate and well-calibrated credit risk estimates.

Supply Chain and Risk Management: An empirical approach in food chain businesses

Departing on our limited understanding of how risks give rise to management controls and performance, this paper aims to better understand the role of Performance Measurement Systems (PMS) and Risk Management Systems (RMS) in food chain businesses. To do so, both case-based research and survey methods are used to develop a comprehensive inventory of the main risks that food supply food managers face, and to provide insight regarding the management control mechanisms they use for enhancing relational performance. Results shows a trend toward a higher management control in food supply relationship, with positive effects on partners’ organisational fit and performance. However some risk sources are still under-managed, as those related to technical uncertainties (transportation problems, long-distance, uncertain technology) and second-tier problems, which represent upcoming challenges in food supply networks. DOI: http://dx.doi.org/10.4995/ICBM.2015.1264