Enhancing Stock Forecasting and Investment Decisions with Stacking Ensemble on Historical Transaction Data

Stock prices are volatile and cannot be predicted accurately, so it requires a model that can capture market dynamics. Investors face the risk of capital loss due to unexpected price changes. Therefore, it is important to develop methods that can help predict stock price movements better. Stock price prediction is one of the important strategies in stock investment to optimize profits. However, this process is challenging because it relies heavily on good quality data sets. A good quality dataset and processed using a stacking ensemble model can increase the high prediction rate. The purpose of this study is to provide more effective tools for investors to make smart investment decisions and reduce the risk of losses in the stock market. The research methodology uses the Stacking Ensemble to predict stock prices and is evaluated using three scoring metrics. The results of the study show that it has obtained an accuracy value from BBCA R2 stocks 0.9997, MAE 0.0093, RMSE 0.0003. This shows that the model is very effective in predicting stock prices.