Optimistic Concurrency Control for Distributed Unsupervised Learning
This addresses the challenge of balancing strict and lax concurrency constraints for large-scale machine learning algorithms, particularly in unsupervised settings, though it appears incremental as it offers an intermediate alternative rather than a breakthrough.
The paper tackles the problem of distributed unsupervised learning by proposing an optimistic concurrency control paradigm that assumes conflicts are rare and resolves them when they occur, demonstrating it in clustering, feature learning, and online facility location with large-scale experiments.
Research on distributed machine learning algorithms has focused primarily on one of two extremes - algorithms that obey strict concurrency constraints or algorithms that obey few or no such constraints. We consider an intermediate alternative in which algorithms optimistically assume that conflicts are unlikely and if conflicts do arise a conflict-resolution protocol is invoked. We view this "optimistic concurrency control" paradigm as particularly appropriate for large-scale machine learning algorithms, particularly in the unsupervised setting. We demonstrate our approach in three problem areas: clustering, feature learning and online facility location. We evaluate our methods via large-scale experiments in a cluster computing environment.