Korea Credit Information Services, Research Results on AI-Based Bankruptcy Rate Prediction by Industry for Sole Proprietors

[Asia Economy Reporter Kim Hyo-jin] The Korea Credit Information Services (KCIS) announced on the 24th that it has achieved results in a study predicting the default rates of individual business owners by industry using artificial intelligence (AI) techniques.

Individual business owners possess attributes of both companies and individuals (representatives), but until now, research reflecting the characteristics of individual business owners has been difficult due to insufficient data.

Accordingly, KCIS conducted this study to develop a model that predicts future default rates by industry for individual business owners by comprehensively analyzing macroeconomic indicators, various corporate credit information, and personal credit information of representatives that it holds.

KCIS explained that through this study, it confirmed that when micro-level credit information held by KCIS is additionally used along with traditionally widely used macroeconomic variables in predicting corporate default rates, the predictive power of default rates improves.

KCIS stated that the AI techniques and development methodology used in this model were presented at the international deep learning expert conference 'DLDC' at the end of last October, attracting attention.

Shin Hyun-jun, President of KCIS, said, "If we can accurately predict the default rates of individual business owners by industry through AI techniques, it will greatly help prevent risks in the economic system," and added, "We will open the meaningful data verified through this study via the 'Big Data Open System' to support financial companies and others in utilizing it as a dataset for AI development."