Partial Distribution Alignment via Adaptive Optimal Transport
This addresses the challenge of partial distribution alignment in machine learning, particularly for domain adaptation, by introducing a novel mathematical framework to handle noisy data, though it appears incremental in its application.
The authors tackled the problem of aligning source and target distributions in the presence of noise and outliers by proposing adaptive optimal transport, which allows for adaptive-mass preserving instead of fixed constraints. The method significantly outperformed state-of-the-art algorithms on domain adaptation benchmarks.
To remedy the drawbacks of full-mass or fixed-mass constraints in classical optimal transport, we propose adaptive optimal transport which is distinctive from the classical optimal transport in its ability of adaptive-mass preserving. It aims to answer the mathematical problem of how to transport the probability mass adaptively between probability distributions, which is a fundamental topic in various areas of artificial intelligence. Adaptive optimal transport is able to transfer mass adaptively in the light of the intrinsic structure of the problem itself. The theoretical results shed light on the adaptive mechanism of mass transportation. Furthermore, we instantiate the adaptive optimal transport in machine learning application to align source and target distributions partially and adaptively by respecting the ubiquity of noises, outliers, and distribution shifts in the data. The experiment results on the domain adaptation benchmarks show that the proposed method significantly outperforms the state-of-the-art algorithms.