用于预测模型可复用性的聚合表示度量
Aggregate Representation Measure for Predictive Model Reusability
AI总结：
本研究提出聚合表示度量ARM，通过量化模型表示在分布偏移下的变化，提前预测重训练所需的训练轮数、能耗与碳排放，助力低成本模型复用及最优架构选择。
摘要：
本文提出一种预测量化指标，用于估计已训练模型在分布偏移下的重训练成本。所提出的聚合表示度量（Aggregated Representation Measure，ARM）量化模型表示从旧数据分布到新数据分布的变化。它能在实际重训练模型之前，给出重训练所需资源——训练轮数（epochs）、能耗与碳排放——的单一简洁指标。这使得复用模型的成本远低于从头训练新模型。实验结果表明，ARM能合理预测不同噪声强度下的重训练成本，还可用于多种模型架构之间的比较，以选出最具成本效益且最可持续的方案。
英文摘要：
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 than training a new model from scratch. The experimental results indicate that ARM reasonably predicts retraining costs for varying noise intensities and enables comparisons among multiple model architectures to determine the most cost-effective and sustainable option.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
