Poster
On the Complexity of Learning Sparse Functions with Statistical and Gradient Queries
Nirmit Joshi · Theodor Misiakiewicz · Nati Srebro
[
Abstract
]
Wed 11 Dec 4:30 p.m. PST
— 7:30 p.m. PST
Abstract:
The goal of this paper is to investigate the complexity of gradient algorithms when learning sparse functions (juntas). We introduce a type of Statistical Queries ($\mathsf{SQ}$), which we call Differentiable Learning Queries ($\mathsf{DLQ}$), to model gradient queries on a specified loss with respect to an arbitrary model. We provide a tight characterization of the query complexity of $\mathsf{DLQ}$ for learning the support of a sparse function over generic product distributions in terms of a certain ``leap exponent'' of the target function (or distribution more generally). This complexity crucially depends on the loss function. For the squared loss, $\mathsf{DLQ}$ matches the complexity of Correlation Statistical Queries $(\mathsf{CSQ})$—potentially much worse than $\mathsf{SQ}$. On the other hand, for simple `generic’ loss functions such as the $\ell_1$ loss, $\mathsf{DLQ}$ always achieves the same complexity as $\mathsf{SQ}$. Finally, we provide evidence that $\mathsf{DLQ}$ can indeed capture learning with (stochastic) gradient descent by showing it correctly describes the complexity of learning with a two-layer neural network in the mean field regime and linear scaling.
Live content is unavailable. Log in and register to view live content