Comparing Oversampling Techniques to Handle the Class Imbalance Problem: A Customer Churn Prediction Case Study

Abstract

Customer retention is a major issue for various service-based organizations particularly telecom industry, wherein predictive models for observing the behavior of customers are one of the great instruments in customer retention process and inferring the future behavior of the customers. However, the performances of predictive models are greatly affected when the real-world data set is highly imbalanced. A data set is called imbalanced if the samples size from one class is very much smaller or larger than the other classes. The most commonly used technique is over/under sampling for handling the class-imbalance problem (CIP) in various domains. In this paper, we survey six well-known sampling techniques and compare the performances of these key techniques, i.e., mega-trend diffusion function (MTDF), synthetic minority oversampling technique, adaptive synthetic sampling approach, couples top-N reverse k-nearest neighbor, majority weighted minority oversampling technique, and immune centroids oversampling technique. Moreover, this paper also reveals the evaluation of four rules-generation algorithms (the learning from example module, version 2 (LEM2), covering, exhaustive, and genetic algorithms) using publicly available data sets. The empirical results demonstrate that the overall predictive performance of MTDF and rules-generation based on genetic algorithms performed the best as compared with the rest of the evaluated oversampling methods and rule-generation algorithms

Predicting the cytotoxicity of chemicals using ensemble learning methods and molecular fingerprints

“…Significantly, this problem often appears in cytotoxic datasets, which makes it difficult for computers to learn fully the features of cytotoxic compounds. As a result, an error will occur when computers classify unknown data (Amin et al, 2016;Estabrooks, 2010;Krawczyk, 2016;Nanni, Fantozzi, & Lazzarini, 2015). This has caused great difficulties in computer-based prediction of toxicity, impeding many studies and preventing application to drug screening.…”

Unsupervised Machine Learning for Networking: Techniques, Applications and Research Challenges

“…Data quality issues must be carefully considered since any problem with the data quality will seriously mar the performance of ML algorithms. A potential problem is that dataset may be imbalanced if the samples size from one class is very much smaller or larger than the other classes [321]. In such imbalanced datasets, the algorithm must be careful not to ignore the rare class by assuming it to be noise.…”

Can We Predict the Scenic Beauty of Locations from Geo-tagged Flickr Images?

“…Similarly we have balanced out the classes in the Paris dataset. Although any other oversampling techniques could have been chosen and they are reported to have competitive performances [1] on various datasets, to avoid complexity we have chosen to apply SMOTE. The oversampled dataset was used to train classifiers and the result of such classifiers can be found in Table 3 and 4.…”