Evaluating the Impact of Telecom Payment Behavior and Equb Participation on Credit Approval Decisions

dc.contributor.advisorRosa Tsegay
dc.contributor.authorSelamawit Getasew
dc.date.accessioned2026-08-04T21:22:13Z
dc.date.available2026-08-04T21:22:13Z
dc.date.issued2026-06
dc.description.abstractAvailability of credit helps in the development of small-scale businesses, and in most cases, traditional credit scoring is not effective for people without adequate financial history. This study aims to determine whether alternative data helps in improving the accuracy of microfinance institutions in predicting the success of loans. This study aims to determine the effect of using alternative data in improving the accuracy of microfinance institutions in predicting the success of loans. This is done by comparing the performance of models using traditional data and models using traditional and alternative data. The data preprocessing steps involve handling missing values, limiting the number of outliers, scaling, and encoding. Stratified sampling and 5-fold cross-validation are used to perform the study. The performance is measured by accuracy, precision, recall, and F1 score. The results obtained from the study show marginal improvements in the accuracy of microfinance institutions in predicting the success of loans. LightGBM secures the highest result, with an average F1 score of 0.857 that is realized when conventional features are employed. Other models’ (for instance, Random Forest and XGBoost) effectiveness is at the par with LightGBM. Incorporating alternative data hardly changes the models’ effectiveness. The addition of alternative data to traditional data helps in enhancing the accuracy of microfinance institutions in predicting the success of loans, although the improvements are marginal and context-dependent
dc.identifier.urihttps://etd.aau.edu.et/handle/123456789/8804
dc.language.isoen
dc.publisherAddis Ababa University
dc.subjectArtificial intelligence
dc.subjectalternative data
dc.subjectinclusive lending
dc.subjectcredit evaluation
dc.subjectsmall business finance
dc.subjectpredictive modelling
dc.subjectmachine learning
dc.subjecteconomic empowerment
dc.subjectequitable finance
dc.titleEvaluating the Impact of Telecom Payment Behavior and Equb Participation on Credit Approval Decisions
dc.typeThesis

Files

Original bundle
Now showing 1 - 1 of 1
No Thumbnail Available
Name:
Selamawit_ Getasew_ 2026.pdf
Size:
1 MB
Format:
Adobe Portable Document Format
License bundle
Now showing 1 - 1 of 1
No Thumbnail Available
Name:
license.txt
Size:
1.71 KB
Format:
Item-specific license agreed to upon submission
Description: