DJIA (2009~2023) & KOSPI50 (2016~2023) Dataset

This dataset provides curated financial time series data from two major equity indices: the Dow Jones Industrial Average (DJIA) in the United States and the KOSPI 50 in South Korea. It includes per-asset and market-level features, historical price-based input sequences, asset returns, and return-based asset rankings, all formatted to support machine learning research in quantitative finance. The data covers both training (2009\u20132021 for DJIA, 2016\u20132021 for KOSPI) and testing periods (2022\u20132023). All price features are normalized by the previous trading day's values to ensure scale invariance across assets. This dataset is particularly suited for applications such as stock ranking, portfolio optimization via reinforcement learning, and time-series forecasting using sequence models.