Currently, adaptive strategies that classify micro problems as active or inactive rely on a similarity metric, whereas model switching uses a user-defined switching function. These two functionalities can be used together: the adaptive strategy classifies problems as active or inactive, and instead of not solving inactive problems, reduced models can be solved to obtain their solutions. If there are more than two models on the micro scale, an algorithm would need to be created to determine when to switch to the third level.
Currently, adaptive strategies that classify micro problems as active or inactive rely on a similarity metric, whereas model switching uses a user-defined switching function. These two functionalities can be used together: the adaptive strategy classifies problems as active or inactive, and instead of not solving inactive problems, reduced models can be solved to obtain their solutions. If there are more than two models on the micro scale, an algorithm would need to be created to determine when to switch to the third level.