Reframing in context: A systematic approach for model reuse in machine learning

Automating data science

“…Once models are considered unfit or degraded, retraining to some new data that has shifted from the original data seems easily mechanizable (repeat-ing the experiment), but it depends on whether the operating conditions that were used initially still hold after the data shift. Reliable and well understood models can often be reused even in new or changing circumstances, through domain adaptation, transfer learning, lifelong learning, or reframing; 20 this represents a more compositional form of automation.…”

Automating Data Science: Prospects and Challenges

“…Once models are considered unfit or degraded, retraining to some new data that has shifted from the original data seems easily mechanizable (repeating the experiment), but it depends on whether the operating conditions that were used initially still hold after the data shift. Reliable and wellunderstood models can often be reused even in new or changing circumstances, through domain adaptation, transfer learning, lifelong learning, or reframing [20]; this represents a more compositional form of automation.…”

Batchwise Patching of Classifiers

“…In this respect it is also related to boosting or additive logistic regression (Friedman, Hastie, and Tibshirani 2000), but the setting differs in that there, multiple iterations are performed on the same data, whereas the goal here is the adaptation to new data. It is also similar to reframing (Hernández-Orallo et al 2016), where the goal is to have a unspecific general model that can be specialized for a particular task. Finally, patching may also be viewed as an instance of exceptional model mining (Duivesteijn, Feelders, and Knobbe 2016), in that it also focuses on recognizing and modeling differences to a base model.…”