DRIVE: One-bit Distributed Mean Estimation
This addresses communication efficiency in distributed and federated learning, but appears incremental as it builds on existing compression techniques.
The paper tackles the problem of compressing d-dimensional vectors from n clients into d(1+o(1)) bits each to approximately reconstruct their mean, with applications in distributed and federated learning, and shows consistent improvements over state-of-the-art methods in evaluations across various datasets.
We consider the problem where $n$ clients transmit $d$-dimensional real-valued vectors using $d(1+o(1))$ bits each, in a manner that allows the receiver to approximately reconstruct their mean. Such compression problems naturally arise in distributed and federated learning. We provide novel mathematical results and derive computationally efficient algorithms that are more accurate than previous compression techniques. We evaluate our methods on a collection of distributed and federated learning tasks, using a variety of datasets, and show a consistent improvement over the state of the art.