Query Complexity of Least Absolute Deviation Regression via Robust Uniform Convergence
This provides foundational insights into robust regression for machine learning practitioners dealing with outliers, though it is incremental in extending known techniques to new loss functions.
The paper tackles the problem of determining the query complexity for least absolute deviation regression, showing that it is Θ(d/ε²) up to logarithmic factors, and extends this to ℓ_p losses for p∈(1,2).
Consider a regression problem where the learner is given a large collection of $d$-dimensional data points, but can only query a small subset of the real-valued labels. How many queries are needed to obtain a $1+ε$ relative error approximation of the optimum? While this problem has been extensively studied for least squares regression, little is known for other losses. An important example is least absolute deviation regression ($\ell_1$ regression) which enjoys superior robustness to outliers compared to least squares. We develop a new framework for analyzing importance sampling methods in regression problems, which enables us to show that the query complexity of least absolute deviation regression is $Θ(d/ε^2)$ up to logarithmic factors. We further extend our techniques to show the first bounds on the query complexity for any $\ell_p$ loss with $p\in(1,2)$. As a key novelty in our analysis, we introduce the notion of robust uniform convergence, which is a new approximation guarantee for the empirical loss. While it is inspired by uniform convergence in statistical learning, our approach additionally incorporates a correction term to avoid unnecessary variance due to outliers. This can be viewed as a new connection between statistical learning theory and variance reduction techniques in stochastic optimization, which should be of independent interest.