On-Demand Sampling: Learning Optimally from Multiple Distributions

arXiv:2210.12529v3 Announce Type: replace
Abstract: Social and real-world considerations such as robustness, fairness, social welfare and multi-agent tradeoffs have given rise to multi-distribution learning paradigms, such as collaborative learning, group distributionally robust optimization, and fair federated learning. In each of these settings, a learner seeks to uniformly minimize its expected loss over $n$ predefined data distributions, while using as few samples as possible. In this paper, we establish the optimal sample complexity of these learning paradigms and give algorithms that …

abstract agent arxiv collaborative cs.cy cs.lg demand distribution fair fairness federated learning loss multi-agent multiple optimization robust robustness sampling social type welfare world