Artificial Intelligence Applied to Sustainable Fund Forecasting: An Ensemble Methods Approach with Embeddings

This study develops a model for forecasting financial series, focusing on the Corporate Sustainability Index (ISE B3) and the ISUS11 fund, using Artificial Intelligence and Machine Learning techniques. The research combines the Kalman filter and the Bi-LSTM neural network using ensemble methods, with the aim of capturing the complexity and volatility of the financial markets. Model 3, which combines both, showed the best performance, with 90% accuracy in estimating tops and valleys. The combination of models proved to be effective in improving the accuracy and robustness of forecasts, highlighting the importance of hybrid approaches.