Aggregate Representation Measure for Predictive Model Reusability

arXiv:2405.09600v1 Announce Type: new
Abstract: In this paper, we propose a predictive quantifier to estimate the retraining cost of a trained model in distribution shifts. The proposed Aggregated Representation Measure (ARM) quantifies the change in the model's representation from the old to new data distribution. It provides, before actually retraining the model, a single concise index of resources - epochs, energy, and carbon emissions - required for the retraining. This enables reuse of a model with a much lower cost …

abstract arm arxiv change cost cs.ai cs.cv cs.cy cs.lg data distribution paper predictive representation retraining type