Fast, Accurate, and Simple Models for Tabular Data via Augmented Distillation
This addresses the deployment inefficiency of AutoML for tabular data, offering a practical solution for users needing fast and accurate models, though it is incremental as it builds on existing distillation and augmentation techniques.
The paper tackles the problem of deploying large, slow AutoML ensembles on tabular data by proposing FAST-DAD, a distillation method that converts complex ensembles into individual models like boosted trees, achieving over 10x faster and more accurate results across 30 datasets.
Automated machine learning (AutoML) can produce complex model ensembles by stacking, bagging, and boosting many individual models like trees, deep networks, and nearest neighbor estimators. While highly accurate, the resulting predictors are large, slow, and opaque as compared to their constituents. To improve the deployment of AutoML on tabular data, we propose FAST-DAD to distill arbitrarily complex ensemble predictors into individual models like boosted trees, random forests, and deep networks. At the heart of our approach is a data augmentation strategy based on Gibbs sampling from a self-attention pseudolikelihood estimator. Across 30 datasets spanning regression and binary/multiclass classification tasks, FAST-DAD distillation produces significantly better individual models than one obtains through standard training on the original data. Our individual distilled models are over 10x faster and more accurate than ensemble predictors produced by AutoML tools like H2O/AutoSklearn.