Kristof Coussement

Churn prediction in subscription services: An application of support vector machines while comparing two parameter-selection techniques

CRM gains increasing importance due to intensive competition and saturated markets. With the purpose of retaining customers, academics as well as practitioners find it crucial to build a churn prediction model that is as accurate as possible. This study applies support vector machines in a newspaper subscription context in order to construct a churn model with a higher predictive performance. Moreover, a comparison is made between two parameter-selection techniques, needed to implement support vector machines. Both techniques are based on grid search and cross-validation. Afterwards, the predictive performance of both kinds of support vector machine models is benchmarked to logistic regression and random forests. Our study shows that support vector machines show good generalization performance when applied to noisy marketing data. Nevertheless, the parameter optimization procedure plays an important role in the predictive performance. We show that only when the optimal parameter-selection procedure is applied, support vector machines outperform traditional logistic regression, whereas random forests outperform both kinds of support vector machines. As a substantive contribution, an overview of the most important churn drivers is given. Unlike ample research, monetary value and frequency do not play an important role in explaining churn in this subscription-services application. Even though most important churn predictors belong to the category of variables describing the subscription, the influence of several client/company-interaction variables cannot be neglected.

Improving customer attrition prediction by integrating emotions from client/company interaction emails and evaluating multiple classifiers

Predicting customer churn with the purpose of retaining customers is a hot topic in academy as well as in todays business environment. Targeting the right customers for a specific retention campaign carries a high priority. This study focuses on two aspects in which churn prediction models could be improved by (i) relying on customer information type diversity and (ii) choosing the best performing classification technique. (i) With the upcoming interest in new media (e.g. blogs, emails,...), client/company interactions are facilitated. Consequently, new types of information are available which generate new opportunities to increase the prediction power of a churn model. This study contributes to the literature by finding evidence that adding emotions expressed in client/company emails increases the predictive performance of an extended RFM churn model. As a substantive contribution, an in-depth study of the impact of the emotionality indicators on churn behavior is done. (ii) This study compares three classification techniques - i.e. Logistic Regression, Support Vector Machines and Random Forests - to distinguish churners from non-churners. This paper shows that Random Forests is a viable opportunity to improve predictive performance compared to Support Vector Machines and Logistic Regression which both exhibit an equal performance.

Ensemble classification based on generalized additive models

Generalized additive models (GAMs) are a generalization of generalized linear models (GLMs) and constitute a powerful technique which has successfully proven its ability to capture nonlinear relationships between explanatory variables and a response variable in many domains. In this paper, GAMs are proposed as base classifiers for ensemble learning. Three alternative ensemble strategies for binary classification using GAMs as base classifiers are proposed: (i) GAMbag based on Bagging, (ii) GAMrsm based on the Random Subspace Method (RSM), and (iii) GAMens as a combination of both. In an experimental validation performed on 12 data sets from the UCI repository, the proposed algorithms are benchmarked to a single GAM and to decision tree based ensemble classifiers (i.e. RSM, Bagging, Random Forest, and the recently proposed Rotation Forest). From the results a number of conclusions can be drawn. Firstly, the use of an ensemble of GAMs instead of a single GAM always leads to improved prediction performance. Secondly, GAMrsm and GAMens perform comparably, while both versions outperform GAMbag. Finally, the value of using GAMs as base classifiers in an ensemble instead of standard decision trees is demonstrated. GAMbag demonstrates performance comparable to ordinary Bagging. Moreover, GAMrsm and GAMens outperform RSM and Bagging, while these two GAM ensemble variations perform comparably to Random Forest and Rotation Forest. Sensitivity analyses are included for the number of member classifiers in the ensemble, the number of variables included in a random feature subspace and the number of degrees of freedom for GAM spline estimation.

A Bayesian approach for incorporating expert opinions into decision support systems

Interest in the use of (big) company data and data-mining models to guide decisions exploded in recent years. In many domains there are human experts whose knowledge is essential in building, interpreting and applying these models. However, the impact of integrating expert opinions into the decision-making process has not been sufficiently investigated. This research gap deserves attention because the triangulation of information sources is critical for the success of analytical projects. This paper contributes to the decision-making literature by (a) detailing the natural advantages of the Bayesian framework for fusing multiple information sources into one decision support system (DSS), (b) confirming the necessity for adjusted methods in this data-explosion era, and (c) opening the path to future applications of Bayesian DSSs in other organizational research contexts. In concrete, we propose a Bayesian decision support framework that formally fuses subjective human expert opinions with more objective organizational information. We empirically test the proposed Bayesian fusion approach in the context of a customer-satisfaction prediction study and show how it improves the prediction performance of the human experts and a data-mining model ignoring expert information. This paper introduces a decision support framework to fuse information sources.Fusing big data with human opinions ensures higher-quality decisions.The paper demonstrates the advantage of the Bayesian machinery for information fusion.

A comparative analysis of data preparation algorithms for customer churn prediction

Data preparation is a process that aims to convert independent (categorical and continuous) variables into a form appropriate for further analysis. We examine data-preparation alternatives to enhance the prediction performance for the commonly-used logit model. This study, conducted in a churn prediction modeling context, benchmarks an optimized logit model against eight state-of-the-art data mining techniques that use standard input data, including real-world cross-sectional data from a large European telecommunication provider. The results lead to following conclusions. (i) Analysts better acknowledge that the data-preparation technique they choose actually affects churn prediction performance; we find improvements of up to 14.5% in the area under the receiving operating characteristics curve and 34% in the top decile lift. (ii) The enhanced logistic regression also is competitive with more advanced single and ensemble data mining algorithms. This article concludes with some managerial implications and suggestions for further research, including evidence of the generalizability of the results for other business settings. We study the impact of data preparation on customer churn prediction performance.Effective data preparation improves AUC up to 14.5% and top decile lift up to 34%.Optimized logistic regression is competitive with advanced data mining algorithms.

Improving customer retention management through cost-sensitive learning

Purpose – Retailers realize that customer churn detection is a critical success factor. However, no research study has taken into consideration that misclassifying a customer as a non-churner (i.e. predicting that (s)he will not leave the company, while in reality (s)he does) results in higher costs than predicting that a staying customer will churn. The aim of this paper is to examine the prediction performance of various cost-sensitive methodologies (direct minimum expected cost (DMECC), metacost, thresholding and weighting) that incorporate these different costs of misclassifying customers in predicting churn. Design/methodology/approach – Cost-sensitive methodologies are benchmarked on six real-life churn datasets from the retail industry. Findings – This article argues that total misclassification cost, as a churn prediction evaluation measure, is crucial as input for optimizing consumer decision making. The practical classification threshold of 0.5 for churn probabilities (i.e. when the churn probab...

Improving direct mail targeting through customer response modeling

Data-mining algorithms (CHAID, CART and neural networks) perform best.Logistic regression and linear discriminant analysis are statistical alternatives.Quadratic discriminant analysis, naive Bayes, C4.5 and k-NN algorithm perform badly. Direct marketing is an important tool in the promotion mix of companies, amongst which direct mailing is crucial. One approach to improve direct mail targeting is response modeling, i.e. a predictive modeling approach that assigns future response probabilities to customers based on their history with the company. The contributions to the response modeling literature are three-fold. First, we introduce well-known statistical and data-mining classification techniques (logistic regression, linear and quadratic discriminant analysis, naive Bayes, neural networks, decision trees, including CHAID, CART and C4.5, and the k-NN algorithm) to the direct marketing community. Second, we run a predictive benchmarking study using the above classifiers on four real-life direct marketing datasets. The 10-fold cross-validated area under the receiver operating characteristics curve is used as evaluation metric. Third, we give managerial insights that facilitate the classifier choice based on the trade-off between interpretability and predictive performance of the classifier. The findings of the benchmark study show that data-mining algorithms (CHAID, CART and neural networks) perform well on this test bed, followed by simplistic statistical classifiers like logistic regression and linear discriminant analysis. It is shown that quadratic discriminant analysis, naive Bayes, C4.5 and the k-NN algorithm yield poor performance.

A new hybrid classification algorithm for customer churn prediction based on logistic regression and decision trees

Abstract Decision trees and logistic regression are two very popular algorithms in customer churn prediction with strong predictive performance and good comprehensibility. Despite these strengths, decision trees tend to have problems to handle linear relations between variables and logistic regression has difficulties with interaction effects between variables. Therefore a new hybrid algorithm, the logit leaf model (LLM), is proposed to better classify data. The idea behind the LLM is that different models constructed on segments of the data rather than on the entire dataset lead to better predictive performance while maintaining the comprehensibility from the models constructed in the leaves. The LLM consists of two stages: a segmentation phase and a prediction phase. In the first stage customer segments are identified using decision rules and in the second stage a model is created for every leaf of this tree. This new hybrid approach is benchmarked against decision trees, logistic regression, random forests and logistic model trees with regards to the predictive performance and comprehensibility. The area under the receiver operating characteristics curve (AUC) and top decile lift (TDL) are used to measure the predictive performance for which LLM scores significantly better than its building blocks logistic regression and decision trees and performs at least as well as more advanced ensemble methods random forests and logistic model trees. Comprehensibility is addressed by a case study for which we observe some key benefits using the LLM compared to using decision trees or logistic regression.

Maximize What Matters: Predicting Customer Churn With Decision-Centric Ensemble Selection.

Churn modeling is important to sustain profitable customer relationships in saturated consumer markets. A churn model predicts the likelihood of customer defection. This helps to target retention offers to the right customers and use marketing resources efficiently. Several statistical prediction methods exist in marketing, but all these suffer an important limitation: they do not allow the analyst to account for campaign planning objectives and constraints during model building. Our key proposition is that creating churn models in awareness of actual business requirements increases the performance of the final model for marketing decision support. To demonstrate this, we propose a decision-centric framework to create churn models. We test our modeling framework on eight real-life churn data sets and find that it performs significantly better than state-of-the-art churn models. We estimate that our approach increases the per customer profits of retention campaigns by

A framework for configuring collaborative filtering-based recommendations derived from purchase data

.47 on average. Further analysis confirms that this improvement comes directly from maximizing business objectives during model building. The main implication of our study is thus that companies better shift from a purely statistical to a more business-driven modeling approach when predicting customer churn.