论文标题
学习教授公平意识 - 深度多任务学习
Learning to Teach Fairness-aware Deep Multi-task Learning
论文作者
论文摘要
公平感知的学习主要集中于单个任务学习（STL）。多任务学习（MTL）的公平含义直到最近才被考虑，并提出了一种开创性的方法，该方法考虑了每项任务的公平 - 准确性权衡以及不同任务之间的绩效权衡。我们提出了一种灵活的方法，而不是刚性公平 - 准确性的权衡表述，该方法通过选择哪个目标（准确性或公平性）来在每个步骤中进行优化。我们介绍了L2T-FMT算法，该算法是经过协作培训的教师学习网络。学生学会解决公平的MTL问题，而教师指示学生从准确性或公平性中学习，具体取决于每个任务更难学习的内容。此外，每个任务的每个步骤都使用该目标的动态选择将权衡权重从2T减少到T，其中T是任务的数量。我们在三个真实数据集上进行的实验表明，L2T-FMT在最先进的方法上都提高了公平性（12-19％）和准确性（最高2％）。
Fairness-aware learning mainly focuses on single task learning (STL). The fairness implications of multi-task learning (MTL) have only recently been considered and a seminal approach has been proposed that considers the fairness-accuracy trade-off for each task and the performance trade-off among different tasks. Instead of a rigid fairness-accuracy trade-off formulation, we propose a flexible approach that learns how to be fair in a MTL setting by selecting which objective (accuracy or fairness) to optimize at each step. We introduce the L2T-FMT algorithm that is a teacher-student network trained collaboratively; the student learns to solve the fair MTL problem while the teacher instructs the student to learn from either accuracy or fairness, depending on what is harder to learn for each task. Moreover, this dynamic selection of which objective to use at each step for each task reduces the number of trade-off weights from 2T to T, where T is the number of tasks. Our experiments on three real datasets show that L2T-FMT improves on both fairness (12-19%) and accuracy (up to 2%) over state-of-the-art approaches.
