A recent review on optimisation methods applied to credit scoring models

Purpose This paper aims to present a literature review of the most recent optimisation methods applied Credit Scoring Models (CSMs). Design/methodology/approach The research methodology employed technical procedures based on bibliographic and exploratory analyses. A traditional investigation was carried out using Scopus, ScienceDirect Web Science databases. papers selection classification took place in three steps considering only studies English language published electronic journals (from 2008 2022). led up 46 publications (10 presenting reviews 36 proposing CSMs). Findings findings showed that CSMs are usually formulated Financial Analysis, Machine Learning, Statistical Techniques, Operational Research Data Mining Algorithms. main databases used by researchers were banks University California, Irvine. analyses identified 48 CSMs, ones being: Logistic Regression (13%), Naive Bayes (10%) Artificial Neural Networks (7%). authors conclude advances credit score will require new hybrid approaches capable integrating Big Deep Learning algorithms into CSMs. These should have practical issues considered consider for improving level adaptation performance demanded Practical implications results this study might provide considerable application As it aimed demonstrate methods, is highly legal ethical be better adapted It also suggested improvement focused micro small companies sales instalment plans commercial through or Originality/value economic reality surrounding granting has made risk management complex decision-making issue increasingly supported Therefore, satisfies an important gap analysis contribution consists evolution state art future trends at