Scenario
A scenario in Alviss AI is a user-defined set of variable values for testing hypothetical conditions in simulations and predictions.
A scenario in Alviss AI refers to a user-defined set of conditions or variables used in simulations and predictions to model hypothetical situations and forecast their potential outcomes. It allows you to test "what-if" strategies by adjusting factors such as pricing, distribution, media investments, or other commercial drivers, and compare them against a baseline (e.g., current or zero-investment conditions).
Key Aspects
- Purpose: Scenarios help evaluate the impact of changes on key performance indicators (KPIs) like sales, churn, or revenue, enabling data-driven decision-making and strategy optimization.
- Usage in Features:
- Simulations: Define a scenario (e.g., increasing media spend) and contrast it with a baseline to analyze differences in results, such as ROI or performance uplift.
- Predictions: Set specific values for variables over time to generate forecasts, supporting demand planning or sales optimization.
- Benefits: By exploring multiple scenarios, you can identify synergistic opportunities, reduce risks, and maximize business growth—potentially achieving 10-20% sales increases through tweaks like pricing adjustments.
For practical implementation, refer to the Simulations and Predictions sections in the documentation. Use the platform's filtering and visualization tools to refine and interpret scenario results effectively.
Root Mean Squared Error
Root Mean Squared Error (RMSE) measures average prediction error in the same units as the target variable by taking the square root of Mean Squared Error, making it highly interpretable.
Weighted Mean Absolute Percentage Error
Weighted Mean Absolute Percentage Error (WMAPE) improves MAPE by weighting errors proportional to actual values, giving more influence to higher-magnitude observations in skewed or imbalanced data.