Related Experiment Video
Updated: Jul 16, 2026

Following the Dynamics of Structural Variants in Experimentally Evolved Populations
Published on: February 3, 2023
Precursors of extreme increments
Sarah Hallerberg1, Eduardo G Altmann, Detlef Holstein
1Max Planck Institute for the Physics of Complex Systems, Nöthnitzer Str. 38, D-01187 Dresden, Germany.
Predicting extreme time series increments is improved by larger event sizes and weaker correlations. This study analyzes prediction quality using receiver-operator characteristics (ROC) curves for better forecasting.
Area of Science:
- Time series analysis
- Predictive modeling
- Statistical forecasting
Background:
- Extreme increments in time series pose challenges for accurate prediction.
- Understanding precursor dynamics is crucial for forecasting sudden changes.
Purpose of the Study:
- To investigate precursors and predictability of extreme increments in time series.
- To analyze how prediction quality depends on precursor strategy, event size, and correlation strength.
Main Methods:
- Analytical study in autoregressive processes of order 1.
- Numerical simulations using wind speed recordings and long-range correlated processes.
- Evaluation of prediction success using receiver-operator characteristics (ROC) curves.
Main Results:
- Prediction quality increases with larger event sizes.
- Prediction quality increases with decreasing correlation strength.
- Likelihood ratio serves as a summary index for smooth ROC curves.
Conclusions:
- The predictability of extreme time series increments is significantly influenced by event size and correlation.
- Optimal precursor selection and understanding correlation dynamics enhance forecasting accuracy.
Related Concept Videos
Absolute and Local Extreme Values
Exponential Growth
Orders of Magnitude
The order of magnitude is simply a way of rounding numbers consistently to the nearest power of 10. This makes doing rough mental math...
Elevation of Intermediate Points on Vertical Curves
Increasing Function
Higher Derivatives

