Metric-Optimized Example Weights
This method addresses the challenge of handling non-i.i.d. data and black-box metrics in ML applications, though it is incremental as it builds on cost-weighted learning.
The paper tackles the problem of optimizing complex test metrics under distribution shift by learning example weights for a weighted loss function, achieving improved performance on benchmark datasets and real-world applications.
Real-world machine learning applications often have complex test metrics, and may have training and test data that are not identically distributed. Motivated by known connections between complex test metrics and cost-weighted learning, we propose addressing these issues by using a weighted loss function with a standard loss, where the weights on the training examples are learned to optimize the test metric on a validation set. These metric-optimized example weights can be learned for any test metric, including black box and customized ones for specific applications. We illustrate the performance of the proposed method on diverse public benchmark datasets and real-world applications. We also provide a generalization bound for the method.