A Fairness-Aware and Bias-Resilient XAI Framework for Equitable Financial Decision-Making
Bibliographic record
Abstract
Loan-approval prediction is typically considered while auditing a single protected attribute at a time (race, ethnicity, sex, or age).Fairness-Aware, Interpretable, Resilient, and Equitable (FAIRE) is a multi-stage pipeline that combines data-level balancing, in-training debiasing, and post-processing thresholding, which is complemented by global and local explainability and continuous fairness monitoring with a drift trigger.The evaluation spans centralized and federated training with privacy-preserving aggregation using Home Mortgage Disclosure Act (HMDA) loan-level data.At the selected operating point, fairness improves substantially: Demographic Parity (DP) rises from 0.74 [0.72, 0.76] to 0.92 [0.90, 0.94]; the Equal Opportunity (EO) gap declines to 0.05 [0.04, 0.06]; and Equalized Odds (EOdds) decreases to 0.07 [0.06, 0.09].The change in Area under the curve-Receiver-operating characteristic curve (AUC-ROC) changes by 0.5 percentage points relative to the best utility setting.In the federated regime (50 clients, Non-Independent and Identically Distributed (non-IID) partitions), AUC-ROC remains within 1 percentage point of centralized utility, while fairness remains close to centralized post-mitigation levels (e.g., DP 0.90 [0.88, 0.92], EO 0.06 [0.05, 0.07], EOdds 0.11 [0.10, 0.12]), with wider intervals for clients with small protected-group support sample sizes.A composite Interpretability Score increases through higher surrogate fidelity, sparser reason sets, and more stable attributions; SHapley Additive exPlanations (SHAP), Local Interpretable Modelagnostic Explanations (LIME), and Integrated Gradients produce adverse-action-ready reason codes consistent with threshold-style explanations.The resulting pipeline delivers measurable fairness gains with minimal utility cost across centralized and federated settings while maintaining transparent, monitorable credit decisions.
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 | 0.002 | 0.002 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.000 | 0.001 |
| Science and technology studies | 0.001 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.000 | 0.001 |
| Insufficient payload (model declined to judge) | 0.000 | 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".