mP Simulator
The true quality of a demand forecast can best be determined by comparing its output with actual metered values (e.g., settlement data). The challenge, therefore, is to fully evaluate the predictability and prediction accuracy achievable in daily forecast operations by comparison with an earlier period. Most forecasting techniques utilise all available historical data in one modelling session and, whilst this generally provides a useful result, it is not always obvious whether large positive and negative deviations are cancelling each other out or whether on certain days, for yet unknown reasons, there exist even larger deviations. This limitation could be reduced by running many incremental, consecutive modelling sessions and comparing the forecasts against actual values for each day. However, the manual effort required for this process is generally prohibitive. Now, with mP Simulator, the benefits of this, more detailed, modelling can be realised without such manual effort. mP Simulator automates the incremental process allowing consecutive forecasts to be readily calculated (simulated) for an earlier period.
mP Simulator provides important Analysis for Risk Assessment
- Simulates a daily forecast for any historical period
- Detects weaknesses and inconsistencies in the existing data
- Allows quick and simple risk assessment of customers
- Supports the user in improving forecast models
