Forecast Outlier Techniques

Forecast Outlier Techniques are methods used to detect data points in time series or forecasting models that are significantly different from the expected pattern or trend. These outliers may indicate anomalies, errors in data collection, or significant changes in the underlying process being forecasted.

· Some common forecast outlier techniques are:

1. Modified_Z_Score
2. Tukey's_Fences
3. Autoencoder
4. Isolation Forest (IForest)
5. Local Outlier Factor (LOF)