BOK, KSS host forum on AI-driven methods in economic data

BOK, KSS host forum on AI-driven methods in economic data

최초입력 2025.09.19 14:58:50

(BOK)

The Bank of Korea (BOK) and the Korean Statistical Society held a forum on Friday at the BOK annex under the theme “The Evolution of Economic Statistics: AI Applications and Methodological Expansion.”
At the event, Oh Hee-seok, a professor of statistics at Seoul National University, presented on the “Data-Adaptive Factor Model (DAFM),” a method designed to capture underlying patterns by using entire data distribution structures. He noted that the model could more precisely detect idiosyncratic volatility in areas such as stock returns and unemployment forecasts, thus improving prediction accuracy.
In the second session, themed “AI and Big Data-Based Financial Market Forecasting,” scholars discussed new approaches to explaining economic indicators by incorporating diverse sources such as news data, images, and stock market flow trends alongside traditional economic metrics.
Seo Beom-seok, a statistics professor at Sookmyung Women’s University, proposed enhancing exchange rate prediction by employing “multi-view data” that integrates macroeconomic indicators with exchange rate trend images and news data. Han Hee-joo, an economics professor at Sungkyunkwan University, presented methods to improve stock market volatility forecasts using machine learning.
The third session highlighted BOK officials’ efforts to improve existing statistics in response to advances in AI and evolving data needs. These included the development of a news sentiment index generated through small language models (SLMs).
[ⓒ 매일경제 & mk.co.kr, 무단 전재, 재배포 및 AI학습 이용 금지]

매일경제

BOK, KSS host forum on AI-driven methods in economic data

많이 본 뉴스