Investigation of the Application of Machine Learning Algorithms in Credit Risk Assessment of Medium and Micro Enterprises

Abstract

Micro and medium enterprises rely on small finance investments and returns to ensure smooth and prolonged functional processes. Due to improper planning and recovery, credit-based investments and lending sometimes lead to financial downfalls. This article analyzes the adverse impacts of credit risk in such enterprises with a solution. This article introduces a Consistent Assessment Method (CAM) using Linear Learning Analysis (LLA) to address the common issues in credit investments and lending. The proposed method identifies the non-recoverable and prolonged investments that cause overheads over the different financial quarters. Identifying such investments/ lending provides knowledge of credit risks over varying market fluctuations. The learning provides two linear analyses: credit increment and financial development. These analyses are performed quarterly, considering the investment and its return consecutively after the awaited period. The issues are analyzed using linear learning that identifies inconsistent financial variations between investment and credit based on fluctuations and non-recoverable instances. Such identification aids in classifying the variants between financial improvement or risks under different quarters for better financial revisions. Thus, this proposed method achieves 14.65% high-risk detection with 5.56% financial revisions for the industry considered by the data source.

Keywords

• Credit risk
• financial variation
• linear learning
• medium and micro enterprises