Towards Bayesian Data Selection
This provides a theoretical framework for improving data selection in iterative ML algorithms, though it appears incremental as it builds on existing decision theory concepts.
The paper tackles the problem of data selection in iterative machine learning algorithms by framing it as a decision problem to find Bayes-optimal selections, and shows that this approach mitigates confirmation bias in semi-supervised learning with empirical results on simulated and real-world data.
A wide range of machine learning algorithms iteratively add data to the training sample. Examples include semi-supervised learning, active learning, multi-armed bandits, and Bayesian optimization. We embed this kind of data addition into decision theory by framing data selection as a decision problem. This paves the way for finding Bayes-optimal selections of data. For the illustrative case of self-training in semi-supervised learning, we derive the respective Bayes criterion. We further show that deploying this criterion mitigates the issue of confirmation bias by empirically assessing our method for generalized linear models, semi-parametric generalized additive models, and Bayesian neural networks on simulated and real-world data.