Selective Mixup Fine-Tuning for Optimizing Non-Decomposable Objectives

arXiv:2403.18301v1 Announce Type: new
Abstract: The rise in internet usage has led to the generation of massive amounts of data, resulting in the adoption of various supervised and semi-supervised machine learning algorithms, which can effectively utilize the colossal amount of data to train models. However, before deploying these models in the real world, these must be strictly evaluated on performance measures like worst-case recall and satisfy constraints such as fairness. We find that current state-of-the-art empirical techniques offer sub-optimal performance …

abstract adoption algorithms arxiv cs.ai cs.cv cs.lg data fine-tuning however internet machine machine learning machine learning algorithms massive semi-supervised stat.ml supervised machine learning train type usage world