Trade-offs in Large-Scale Distributed Tuplewise Estimation and Learning
This addresses a bottleneck in distributed computing for machine learning practitioners working on non-separable tuplewise problems, offering an incremental improvement over existing methods.
The paper tackles the challenge of scaling tuplewise statistical learning tasks like metric learning and clustering in distributed systems, proposing a repartitioning strategy that balances accuracy and runtime, with theoretical and experimental validation showing variance reduction and efficiency gains.
The development of cluster computing frameworks has allowed practitioners to scale out various statistical estimation and machine learning algorithms with minimal programming effort. This is especially true for machine learning problems whose objective function is nicely separable across individual data points, such as classification and regression. In contrast, statistical learning tasks involving pairs (or more generally tuples) of data points - such as metric learning, clustering or ranking do not lend themselves as easily to data-parallelism and in-memory computing. In this paper, we investigate how to balance between statistical performance and computational efficiency in such distributed tuplewise statistical problems. We first propose a simple strategy based on occasionally repartitioning data across workers between parallel computation stages, where the number of repartitioning steps rules the trade-off between accuracy and runtime. We then present some theoretical results highlighting the benefits brought by the proposed method in terms of variance reduction, and extend our results to design distributed stochastic gradient descent algorithms for tuplewise empirical risk minimization. Our results are supported by numerical experiments in pairwise statistical estimation and learning on synthetic and real-world datasets.