Optimizing Credit Scoring in P2P Lending Using Hybrid Ensemble Learning
Bibliographic record
Abstract
As mentioned in this paper, it is very urgent to improve the credit scoring of peer-to-peer (P2P) lending platform: a credit scoring system based on ensemble learning. Credit scoring systems, which analyze large amounts of data, may have difficulties in processing due to the intrinsic shortcomings of traditional models, namely data imbalance and their dependence on multiple variables. To overcome these difficulties, the proposed methodology uses a hybrid ensemble learning technique, combining the below machine learning algorithms: random forest, gradient boosting, XGBoost and a stacking ensemble model, Hareum algorithm. This method includes improved feature extraction to enhance the learning ability of the model and SMOTE is used to fix the data imbalance between defaulted and non-defaulted loans. Furthermore, SHAP (Shapley Additive Explanation) mean is used to clear the relationship between the choice of features and credit rating in the model. This interpretability is essential for financial applications, as it guarantees system performance and results to end-users. Models are checked according to particular criteria, such as accuracy, precision, recall, F1 score and area under the ROC curve (AUC-ROC). The results show that the ensemble model from stacking achieves better accuracy (92.4%) than the individual models, with an area under the ROC curve (AUC-ROC) of 94.5%. The evaluation of the model showed that the integration of feature engineering and SMOTE significantly improves its performance, especially in terms of recall and accuracy for payment default detection.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.000 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.002 | 0.002 |
| Science and technology studies | 0.001 | 0.000 |
| Scholarly communication | 0.001 | 0.003 |
| Open science | 0.000 | 0.001 |
| Research integrity | 0.000 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".