Towards Fair Disentangled Online Learning for Changing Environments
Towards Fair Disentangled Online Learning for Changing Environments
复制标题
DOI：
10.1145/3580305.3599523
复制
发表时间：
2023-05
期刊：
影响因子：
--
通讯作者：
Chenxu Zhao;Feng Mi;Xintao Wu;Kai Jiang;L. Khan;Christan Earl Grant;Feng Chen
中科院分区：
文献类型：
--
作者：
Chenxu Zhao;Feng Mi;Xintao Wu;Kai Jiang;L. Khan;Christan Earl Grant;Feng Chen
在针对变化环境的在线学习问题中，随着时间的推移，数据被顺序地一个接一个地接收，并且它们的分布假设可能频繁地变化。虽然现有的方法通过提供对动态遗憾或自适应遗憾的严格限制来证明其学习算法的有效性，但它们中的大多数完全忽略了具有模型公平性的学习，模型公平性被定义为跨不同子群体的统计奇偶性（例如，种族和性别）。另一个缺点是，当适应新的环境时，在线学习者需要用全局变化来更新模型参数，这是昂贵且低效的。受稀疏机制转移假设[22]的启发，我们声称在线学习中不断变化的环境可以归因于特定于环境的学习参数的部分变化，其余部分对变化的环境保持不变。为此，在本文中，我们提出了一种新的算法的假设下，在每个时间收集的数据可以解开两个表示，一个环境不变的语义因素和环境特定的变化因素。语义因子进一步用于组公平性约束下的公平预测。为了评估由学习器生成的模型参数序列，提出了一种新的遗憾，其中它采取了混合形式的动态和静态的遗憾指标，其次是公平意识的长期约束。详细的分析为损失后悔和违反累积公平约束提供了理论保证。在真实数据集上的实验结果表明，该方法在模型准确性和公平性方面依次优于基线方法。
In the problem of online learning for changing environments, data are sequentially received one after another over time, and their distribution assumptions may vary frequently. Although existing methods demonstrate the effectiveness of their learning algorithms by providing a tight bound on either dynamic regret or adaptive regret, most of them completely ignore learning with model fairness, defined as the statistical parity across different sub-population (e.g., race and gender). Another drawback is that when adapting to a new environment, an online learner needs to update model parameters with a global change, which is costly and inefficient. Inspired by the sparse mechanism shift hypothesis [22], we claim that changing environments in online learning can be attributed to partial changes in learned parameters that are specific to environments and the rest remain invariant to changing environments. To this end, in this paper, we propose a novel algorithm under the assumption that data collected at each time can be disentangled with two representations, an environment-invariant semantic factor and an environment-specific variation factor. The semantic factor is further used for fair prediction under a group fairness constraint. To evaluate the sequence of model parameters generated by the learner, a novel regret is proposed in which it takes a mixed form of dynamic and static regret metrics followed by a fairness-aware long-term constraint. The detailed analysis provides theoretical guarantees for loss regret and violation of cumulative fairness constraints. Empirical evaluations on real-world datasets demonstrate our proposed method sequentially outperforms baseline methods in model accuracy and fairness.