Double descent for least-squares interpolation on contaminated data: A simulation study
For researchers in robust statistics and overparametrized models, this work shows that double descent can mitigate the impact of data contamination, challenging the need for robust estimators in highly overparametrized regimes.
The study investigates whether double descent occurs in linear regression with contaminated training data, finding that large overparametrization enables double descent, allowing the non-robust least-squares interpolator to outperform robust alternatives in generalization.
Overparametrized models can exhibit an excellent generalization performance, although they should be prone to overfitting according to classical statistical theory. The discovery of the "double descent", indicating that the generalization error decreases after a certain model complexity has been reached, opened a new line of research. Robust statistics considers statistical estimation on contaminated data, which, due to assumptions that do not hold on real data, let data points appear as outliers w.r.t. the assumed "ideal" distribution, potentially severely distorting any classical estimator. We address the question whether a double descent phenomenon can be observed in a linear regression setting with contaminated training data. We compare the performance of the highly non-robust least-squares interpolation estimator with several robust alternatives. It turns out that large overparametrization indeed allows for a double descent phenomenon, resulting in a very good generalization performance of the least-squares interpolator, surpassing that of the robust alternatives.