Bootstrap Bias Corrections for Ensemble Methods
This addresses bias correction for statistical inference in machine learning, particularly for ensemble methods, but appears incremental as it builds on existing bootstrap techniques.
The paper tackles bias in machine learning regression methods by proposing a residual bootstrap for bias correction, demonstrating that it improves bias and predictive accuracy, with test-set accuracy gains of up to 70% over random forests on UCI datasets.
This paper examines the use of a residual bootstrap for bias correction in machine learning regression methods. Accounting for bias is an important obstacle in recent efforts to develop statistical inference for machine learning methods. We demonstrate empirically that the proposed bootstrap bias correction can lead to substantial improvements in both bias and predictive accuracy. In the context of ensembles of trees, we show that this correction can be approximated at only double the cost of training the original ensemble without introducing additional variance. Our method is shown to improve test-set accuracy over random forests by up to 70\% on example problems from the UCI repository.