{"paper_id":"2402.03806","title":"Explainable Automated Machine Learning for Credit Decisions: Enhancing Human Artificial Intelligence Collaboration in Financial Engineering","authors":["Marc Schmitt"],"paper_date":"2024-02-06","paper_type":"empirical","paper_url":"https://arxiv.org/pdf/2402.03806v1","page_url":"https://thequant.space/flowcharts/explainable-automated-machine-learning-for-credit-decisions/","scores":{"empirical_rigor":7,"math_complexity":4,"quadrant":"Street Traders"},"proposes":{"status":"available","statement":"The paper proposes integrating Explainable Automated Machine Learning (AutoML) with SHAP XAI methods to improve transparency, efficiency, and accuracy of credit decision‑making.","quote":"This study demonstrates how the combination of AutoML and XAI not only enhances the efficiency and accuracy of credit decisions but also fosters trust and collaboration between humans and AI systems.","summary":"This paper explores the integration of Explainable Automated Machine Learning (AutoML) in the realm of financial engineering, specifically focusing on its application in credit decision-making. The rapid evolution of Artificial Intelligence (AI) in finance has necessitated a balance between…","strategy":null},"assumptions":{"status":"available","items":[{"assumption":"Random under‑sampling is used to balance the class distribution between good and bad credit applicants.","type":"data","stated":"explicit","quote":"To mitigate bias towards the majority class in this classification study, random under-sampling was employed to balance the dataset."},{"assumption":"Categorical variables are converted to numeric form using one‑hot encoding.","type":"data","stated":"explicit","quote":"categorical values in the data were converted to numerical form using one-hot encoding"},{"assumption":"The publicly available Taiwan and Germany credit datasets accurately reflect information typically accessible to retail banks.","type":"data","stated":"explicit","quote":"These datasets are publicly accessible, facilitating the reproducibility of the empirical analysis."},{"assumption":"The SHAP and variable importance heatmap visualisations are sufficient to provide guidance for human analysts to interpret model behaviour.","type":"model","stated":"explicit","quote":"Run explainability methods on the leaderboard (e.g. Shapley Values, Variable Importance Heatmap)"}]},"datasets":{"status":"available","items":[{"name":"Taiwan credit card dataset","provider":null,"kind":"other","access":"public","period":null,"universe":null,"quote":"Dataset 1 – Taiwan: The first dataset represents payment information from Taiwanese credit card clients. It was first used by Yeh and Lien (2009) and contains 30,000 observations where 6,636 are flagged as defaults."},{"name":"Germany bank credit dataset","provider":null,"kind":"other","access":"public","period":null,"universe":null,"quote":"Dataset 2 – Germany: The second dataset represents detailed customer-level data from a German bank and contains 1,000 observations where 300 are flagged as defaults."}],"known":[],"sample":{"start":null,"end":null,"market":null,"asset_class":null,"frequency":null},"data_resolution":"unclear"},"transaction_costs":{"status":"not_applicable","modeled":"n/a","headline_net_of_costs":null,"detail":null,"quote":null,"source":"front_matter","rule":"R3","candidates":{"front_matter":"n/a","evidence_card":"n/a","evidence_card_quoted":false},"meaning":"the paper reports no trading or execution result"},"results":{"status":"available","evaluation_type":"forecast-accuracy","findings":[{"finding":"Integrating AutoML with SHAP XAI methods yields more transparent and human‑centric credit scoring, bridging the gap between black‑box models and effective human‑AI collaboration.","kind":"other","quote":"The integration of AutoML with explainable AI (XAI) methods, particularly using SHAP value plots and variable importance heatmaps, has illustrated the path towards more transparent, and human‑centric decision‑making in credit scoring."},{"finding":"Explainable AutoML provides visual guidance (SHAP plots, heatmaps) that supports analysts in interpreting model behavior and improves trust in AI‑driven credit decisions.","kind":"other","quote":"Output: a set of visuals to provide guidance for human analysts to interpret model behavior, supporting informed decision-making and model validation."}],"reported_performance":null,"out_of_sample":{"method":"holdout","headline_out_of_sample":null}},"independent_evidence":{"status":"available","supports":[{"paper":{"paper_id":"2306.02773","title":"Explaining AI in Finance: Past, Present, Prospects","paper_date":"2023-06-05","authors":["Barry Quinn"],"empirical_rigor":1,"page_url":"https://thequant.space/flowcharts/explaining-ai-in-finance-past-present-prospects/"},"question":"Do SHAP-based XAI methods increase interpretability and user trust in AI-driven financial decision‑making?","this_paper_finding":"Explainable AutoML provides visual guidance (SHAP plots, heatmaps) that supports analysts in interpreting model behavior and improves trust in AI‑driven credit decisions.","other_paper_finding":"XAI techniques, such as SHAP values, improve interpretability of AI models and increase user trust in financial decision-making.","other_paper_quote":"The paper demonstrates, through simulations, that XAI enhances trust in AI systems, fostering more responsible decision-making within finance.","note":"Both findings report that SHAP visualizations (plots, heatmaps) improve model interpretability and boost trust in AI finance/credit decisions.","confidence":0.756},{"paper":{"paper_id":"2306.02773","title":"Explaining AI in Finance: Past, Present, Prospects","paper_date":"2023-06-05","authors":["Barry Quinn"],"empirical_rigor":1,"page_url":"https://thequant.space/flowcharts/explaining-ai-in-finance-past-present-prospects/"},"question":"How does applying SHAP XAI affect interpretability and trust in AI-driven financial decision-making?","this_paper_finding":"Integrating AutoML with SHAP XAI methods yields more transparent and human‑centric credit scoring, bridging the gap between black‑box models and effective human‑AI collaboration.","other_paper_finding":"XAI techniques, such as SHAP values, improve interpretability of AI models and increase user trust in financial decision-making.","other_paper_quote":"The paper demonstrates, through simulations, that XAI enhances trust in AI systems, fostering more responsible decision-making within finance.","note":"Both findings report that SHAP XAI improves model interpretability and user trust; B extends this by integrating SHAP with AutoML for credit scoring, thereby supporting A’s claim.","confidence":0.683}],"contradicts":[],"qualifies":[],"compared_with":10,"citations":{"cited_by_in_corpus":2,"by_independent_authors":2,"how_cited":{"used_as_baseline":0,"uses_method_or_data":0,"extends":0,"discusses_findings":0,"mentions":2,"unclassified":0},"independent_papers":[],"scope":"Citations between papers in this corpus only; citations from outside it are not counted."}},"implementation":{"status":"available","public":false,"code_url":null,"note":"No repository presented by the authors as their own was found in the paper. Absence here is not proof that no code exists."},"replication":{"status":"none_found","card":null,"note":"No replication of this paper has been run by thequant.space yet."},"extraction":{"text_source":"fulltext","model":"openai/gpt-oss-120b","version":1,"extracted_at":"2026-11-14T18:15:13Z"},"provenance":{"claims":"Everything here is what the authors state in the paper. Nothing was re-run or reproduced by the platform.","quotes":"Assumptions, datasets, cost detail and findings were extracted by a language model. Every quote is verbatim paper text. Where an item has `support`, \"full\" means every figure in it appears in the quoted sentences and \"partial\" means a figure is stated elsewhere in the paper; items with a figure found nowhere in the paper were discarded.","datasets_known":"datasets.known lists well-known vendors and public datasets by a stable id, found by a dictionary over the full text where the sentence says the data is used; it is independent of the model and covers only names in the dictionary.","flags":"transaction_costs.modeled, results.evaluation_type and results.out_of_sample.method are platform classifications (structured-decision models); see https://thequant.space/score-guide/#penalty-flags","comparisons":"supports / contradicts / qualifies are a language model's reading of two papers' quoted findings, restricted to closely related papers that share no author. They are pointers to read, not verified replications, and coverage is partial.","implementation":"code_url is a repository the paper itself presents as its own (rule-scored, ambiguous cases judged by a model), checked for liveness when it was recorded."},"docs":"https://thequant.space/agents/"}