Building an Automated Missing Value Imputation Framework for Multi-Point River Streamflow Data

Puntos clave

• The research aims to build a framework for automatically filling in missing river streamflow data.
• Utilized the MissForest algorithm for missing data imputation.
• Created data by artificially removing 10% of observed values.
• Compared various regression models within the MissForest framework.
• Evaluated imputation accuracy using the coefficient of determination (R²).
• Achieved an average R² of 0.8 with the simplest linear regression model.
• Found that not selecting data for the MissForest process resulted in higher imputation accuracy.
• Demonstrated that MissForest can effectively automate the imputation process for large-scale streamflow data.