University of Twente Student Theses
Bridging the Gap : Enhancing Loan Approval Processes with Machine Learning and Soft Information Integration
Weenink, M.R. (2024) Bridging the Gap : Enhancing Loan Approval Processes with Machine Learning and Soft Information Integration.
|
PDF
419kB |
| Abstract: | The incorporation of machine learning algorithms into financial technology has drastically revolutionized lending procedures by combining soft and hard information. However, the use of soft information such as human characteristics, company prospects, and subjective assessments by loan officers is still underexplored. This study investigates the use of machine learning techniques, notably Support Vector Machines (SVM), to improve loan approval procedures by exploiting soft data. The study uses data from a microfinance organization including both hard and soft information gathered through direct borrower contacts. The data were preprocessed and features were engineered with Term Frequency-Inverse Document Frequency (TF-IDF). Various SVM models with different kernels were trained and assessed for accuracy and the Matthews Correlation Coefficient (MCC), with the model using the linear kernel achieving the highest MCC of 0.726. The findings demonstrate that integrating soft information into SVM models enhances the accuracy of loan disbursement forecasts. Feature extraction revealed that business-related features were the most influential in the model’s decisions, followed by financial and entrepreneur features. A rigorous comparison of actual loan disbursements with model predictions showed that incorporating soft information improves prediction accuracy and reduces repayment difficulties. These findings emphasize the importance of soft information in credit assessments and demonstrate the efficacy of SVM models in real-world lending scenarios, thereby promoting more inclusive and accurate credit evaluation processes. |
| Item Type: | Essay (Bachelor) |
| Faculty: | BMS: Behavioural, Management and Social Sciences |
| Subject: | 85 business administration, organizational science |
| Programme: | International Business Administration BSc (50952) |
| Link to this item: | https://purl.utwente.nl/essays/100270 |
| Export this item as: | BibTeX EndNote HTML Citation Reference Manager |
Show download statistics for this publication
Show download statistics for this publicationRepository Staff Only: item control page