<![CDATA[This study aims to develop an educational software platform that supports Indonesian students in learning stock market behavior by addressing low financial literacy, limited analytical skills, and emotion-driven investment decisions. The proposed platform integrates historical data from liquid IDX80 stocks with a Long Short-Term Memory (LSTM) forecasting model enhanced by technical indicators and calendar-based market features, and presents predicted price movements and trend classifications through an interactive, user-friendly learning interface. Experimental results show that the model achieved strong predictive performance on most IDX80 stocks, demonstrating its ability to capture temporal price patterns, while the inclusion of technical and calendar-based features improved prediction clarity and trend interpretability for student users. Variations in forecasting accuracy across stocks indicate that liquidity and volatility influence model performance, highlighting the importance of contextual interpretation in learning. In conclusion, the findings indicate that integrating LSTM-based forecasting with instructional design principles can support experiential and data-driven investment learning, and the developed platform demonstrates strong potential as both a forecasting tool and an educational technology medium for supporting financial literacy development. ]]>
Copyrights © 2025
