[2K1] [General Session] 13. AI Application

[2K1-04] Analysis of News and Stock Returns in Korean Markets Using High Frequency Data: Exploring Conditions of Classification Algorithm

Sungjae Yoon¹, 〇Aiko Suge¹, Hiroshi Takahashi¹ (1. Keio University)

Keywords:hyperparameter, Long Short Term Memory, high frequency data, text classification

News articles play an important role in financial markets.　This study　analyzes the relationship between news articles and stock price fluctuations using high frequency trading data in Korean stock markets.　Especially, we analyze differences in market reactions according to　languages of news articles. In order to understand the influences of　news articles, this study explores conditions of Long Short Term Memory　(LSTM) models that classify news articles.