Scalable Set Encoding with Universal Mini-Batch Consistency and Unbiased Full Set Gradient Approximation
This work addresses scalability and flexibility issues in set encoding for machine learning applications, offering incremental improvements over prior methods.
The paper tackled the limited expressive power of mini-batch consistent set functions and the inefficiency of computing full set gradients for large sets, proposing a universally MBC class and an efficient training algorithm that achieved constant memory overhead and unbiased gradient approximations across various tasks like image completion and cancer detection.
Recent work on mini-batch consistency (MBC) for set functions has brought attention to the need for sequentially processing and aggregating chunks of a partitioned set while guaranteeing the same output for all partitions. However, existing constraints on MBC architectures lead to models with limited expressive power. Additionally, prior work has not addressed how to deal with large sets during training when the full set gradient is required. To address these issues, we propose a Universally MBC (UMBC) class of set functions which can be used in conjunction with arbitrary non-MBC components while still satisfying MBC, enabling a wider range of function classes to be used in MBC settings. Furthermore, we propose an efficient MBC training algorithm which gives an unbiased approximation of the full set gradient and has a constant memory overhead for any set size for both train- and test-time. We conduct extensive experiments including image completion, text classification, unsupervised clustering, and cancer detection on high-resolution images to verify the efficiency and efficacy of our scalable set encoding framework. Our code is available at github.com/jeffwillette/umbc