Dataset for stock price prediction with LSTM and technical indicators: Evidence from Taiwan-listed firms
Description
This dataset provides daily stock price and trading volume information for three Taiwan-listed firms, covering the period from January 2010 to January 2024. The raw data include open, high, low, and close (OHLC) prices as well as trading volumes, retrieved from the Taiwan Stock Exchange through the Yahoo Finance API. To facilitate financial forecasting research, the dataset also includes derived technical indicators such as moving averages (5-day, 10-day, 20-day), Relative Strength Index (RSI), and Moving Average Convergence Divergence (MACD). In addition, the data are structured into training-ready input–output matrices for Long Short-Term Memory (LSTM) neural networks using a sliding window approach. This dataset is intended for use in stock price prediction, technical trading strategy evaluation, and machine learning research in finance. It can also be applied for replication studies, cross-industry comparisons, and educational purposes.
Files
Steps to reproduce
The dataset was compiled from publicly available stock price data retrieved via the Yahoo Finance API. Technical indicators were calculated using standard financial formulas implemented in Python. The LSTM-ready data matrices were generated using a sliding window approach. Users can reproduce the dataset using similar data sources and standard Python libraries (pandas, TA-lib, NumPy).