Back to Products

Time Series Diagnostics & What-if scenario analysis

Intervention-Aware Modeling

Understanding how localized changes in one variable affect others in multivariate time series is essential for diagnostics and decision-making. Our framework combines temporal decomposition with frequency-domain feature correlation modeling to enable realistic "what-if" scenario analysis.

Key Innovations

  • Realistic propagation across correlated variables
  • Target feature prediction and simulation analysis for updated features

Applications

Time Series Forecasting · Decision Support Systems

Request Information

What-If Feature Simulation: Analyze how modifying an input feature (e.g., DEWP) impacts the target prediction — helping operators understand causal dependencies.

Adjusted Prediction Figure

[figures/adjusted_prediction.png]

Effect of modifying an input feature (DEWP) on the model's prediction. "Adjusted Prediction" illustrates how the target output responds to the modification, highlighting the model's sensitivity to feature-level changes.

Time Series What-if Analysis