Tutorial - applying extreme value theory to characterize food-processing systems
Research output: Contribution to journal › Journal article › Research › peer-review
This tutorial presents extreme value theory (EVT) as an analytical tool in process characterization and shows its potential to describe production performance, eg, across different factories, via reliable estimates of the frequency and scale of extreme events. Two alternative EVT methods are discussed: point over threshold and block maxima. We illustrate the theoretical framework for EVT by process data from two different examples from the food-processing industry. Finally, we discuss limitations, decisions, and possibilities when applying EVT for process data.
|Journal||Journal of Chemometrics|
|Number of pages||12|
|Publication status||Published - 2017|
- Extreme value theory, Food processing, Return-level estimation, Return-time estimation