Exploring data mining techniques and algorithms for predicting customer loyalty and loan default risk scenarios at wisdom microfinance, Addis Ababa, Ethiopia

  • Authors:
  • Getachew Hailemariam;Shawndra Hill;Sintayehu Demissie

  • Affiliations:
  • Addis Ababa University;University of Pennsylvania;Addis Ababa University

  • Venue:
  • Proceedings of the International Conference on Management of Emergent Digital EcoSystems
  • Year:
  • 2012

Quantified Score

Hi-index 0.00

Visualization

Abstract

Microfinance Institutions are established with the mission and goals of serving poor people who lacked financial services from the conventional financial institutions but capable of engaging in small scale economic activities. These institutions are found to be quite relevant in the Ethiopian context and in their current state they are established very recently. To date, there are more than 20 recognized microfinance institutions in Ethiopia. Wisdom Microfinance is one of these accredited institutions. This study was aimed at exploring the potential application of data mining techniques for supporting key operational and strategic decisions of microfinance institutions. Qualitative and quantitative data were gathered and used in the study. The qualitative data revealed facts regarding the information requirement of the company and helped the investigator to understand the business rules. The quantitative data were used for conducting model building experiments. WEKA data mining software was used for model building and analysis. Three sets of classifier algorithms were used in the experiment. It was found out that J48 classifier delivered good result in terms of classifying instances correctly. The study recommends future researchers to experiment model building using multiple algorithms and larger datasets.