Analysis of a Two-Layer Neural Network via Displacement Convexity
This provides theoretical guarantees for global convergence in neural network training, addressing a fundamental challenge in machine learning, though it is incremental as it builds on existing gradient flow theories.
The paper tackles the non-convex optimization problem in learning concave functions using two-layer neural networks with bump-like neurons, proving that gradient descent converges to a Wasserstein gradient flow with displacement convexity, leading to exponential convergence rates as the number of neurons and bump width approach limits.
Fitting a function by using linear combinations of a large number $N$ of `simple' components is one of the most fruitful ideas in statistical learning. This idea lies at the core of a variety of methods, from two-layer neural networks to kernel regression, to boosting. In general, the resulting risk minimization problem is non-convex and is solved by gradient descent or its variants. Unfortunately, little is known about global convergence properties of these approaches. Here we consider the problem of learning a concave function $f$ on a compact convex domain $Ω\subseteq {\mathbb R}^d$, using linear combinations of `bump-like' components (neurons). The parameters to be fitted are the centers of $N$ bumps, and the resulting empirical risk minimization problem is highly non-convex. We prove that, in the limit in which the number of neurons diverges, the evolution of gradient descent converges to a Wasserstein gradient flow in the space of probability distributions over $Ω$. Further, when the bump width $δ$ tends to $0$, this gradient flow has a limit which is a viscous porous medium equation. Remarkably, the cost function optimized by this gradient flow exhibits a special property known as displacement convexity, which implies exponential convergence rates for $N\to\infty$, $δ\to 0$. Surprisingly, this asymptotic theory appears to capture well the behavior for moderate values of $δ, N$. Explaining this phenomenon, and understanding the dependence on $δ,N$ in a quantitative manner remains an outstanding challenge.