Claim · model-written summary
Average forecast MAE was 3.52.
Not reportedTaxonMaterialGeneCompoundOrgan / stageInteractorInteractor roleSecond interactorSecond interactor roleTreatmentComparatorSettingDirectionQualifier
Verbatim
e0“The average Root Mean Square Error (RMSE) and Mean Absolute Error (MAE) values were 4.36 and 3.52, respectively.”
Title
Parameter Selection for a Neural Network Model for Forecasting Plant Pest Population Growth
Akerke Akanova; Nazira Ospanova; Galiya Anarbekova Statistics Optimization & Information Computing 2026. doi:10.19139/soic-2310-5070-3317
Other claims from this paper (13)
- Method / resource Helicoverpa armigera forecasts attained the highest R², above 0.82.
- Method / resource Frankliniella occidentalis forecasts attained the highest R², above 0.82.
- Method / resource Loxostege sticticalis forecasts performed poorest, at R² approximately 0.64.
- Association Forecast accuracy was higher in tomato plots, associated with stable irrigation and protection.
- Association Forecast accuracy was higher in sunflower plots, associated with stable irrigation and protection.
- Method / resource Average forecast RMSE was 4.36.
- Method / resource The optimal two-layer 64/32-neuron ReLU ADAM model yielded MAE 2.9.
- Method / resource The optimal model yielded R² 0.83.
- Method / resource Deeper models overfit.
- Method / resource Simplified models had insufficient accuracy.
- Method / resource Configurations differed significantly, p below 0.001.
- Method / resource Validation demonstrated high robustness with minimal accuracy loss across regions and seasons.
- Recommendation The approach has high practical value for agricultural pest monitoring and forecasting integration.