Optimized Credit Score Prediction via an Ensemble Model and SMOTEENN Integration

Abstract

Credit scores are crucial these days for getting approved for a mortgage, assessing business loans, and managing customer finances. A bank's ability to accurately judge someone's creditworthiness directly impacts their bottom line and operations. For a long time, we have used models like decision trees, logical regression, and neural networks to evaluate credit. However, lately, we have started using more advanced distributed models like LightGBM, XGBoost, and CatBoost. These have improved our credit predictions' accuracy but still have significant limitations. This paper looks at combining multiple models into one robust hybrid model. The goal is to get past the performance bottlenecks of older models. Our proposed approach takes credit score prediction to the next level. We hope to improve precision, F1 score, recall, and AUC metrics. Through a detailed comparison of these elements, we will compare the advantages and disadvantages of our designed Ensemble Model and the current general models. By conducting a detailed comparison of these elements, we will comprehensively check the strengths and weaknesses of our designed Ensemble Model against the currently prevalent models, showcasing our results through an objective analysis.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 13 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.