ESG Factor Integration for Growth Forecasting - Training
This workflow combines company financials with ESG scores, builds a regression model to predict future growth, and then estimates the probability of a slowdown for each company.
First, financial and ESG datasets are joined. The data is split into training (past years) and test (latest year) sets. A Random Forest regression model is trained on the historical data, then used to predict growth for the test set. The model's residuals (errors) are analyzed to estimate how likely each company is to fall below a global slowdown threshold. The results include both point forecasts and slowdown probabilities, which are saved for further analysis.