A Local Convergence Theory for Mildly Over-Parameterized Two-Layer Neural Network

by   Mo Zhou, et al.

While over-parameterization is widely believed to be crucial for the success of optimization for the neural networks, most existing theories on over-parameterization do not fully explain the reason – they either work in the Neural Tangent Kernel regime where neurons don't move much, or require an enormous number of neurons. In practice, when the data is generated using a teacher neural network, even mildly over-parameterized neural networks can achieve 0 loss and recover the directions of teacher neurons. In this paper we develop a local convergence theory for mildly over-parameterized two-layer neural net. We show that as long as the loss is already lower than a threshold (polynomial in relevant parameters), all student neurons in an over-parameterized two-layer neural network will converge to one of teacher neurons, and the loss will go to 0. Our result holds for any number of student neurons as long as it is at least as large as the number of teacher neurons, and our convergence rate is independent of the number of student neurons. A key component of our analysis is the new characterization of local optimization landscape – we show the gradient satisfies a special case of Lojasiewicz property which is different from local strong convexity or PL conditions used in previous work.


page 1

page 2

page 3

page 4


Excess Risk of Two-Layer ReLU Neural Networks in Teacher-Student Settings and its Superiority to Kernel Methods

While deep learning has outperformed other methods for various tasks, th...

Understanding How Over-Parametrization Leads to Acceleration: A case of learning a single teacher neuron

Over-parametrization has become a popular technique in deep learning. It...

Orthogonal Over-Parameterized Training

The inductive bias of a neural network is largely determined by the arch...

Optimization and Generalization of Shallow Neural Networks with Quadratic Activation Functions

We study the dynamics of optimization and the generalization properties ...

The Effects of Mild Over-parameterization on the Optimization Landscape of Shallow ReLU Neural Networks

We study the effects of mild over-parameterization on the optimization l...

Trainability and Data-dependent Initialization of Over-parameterized ReLU Neural Networks

A neural network is said to be over-specified if its representational po...

Optimal Rate of Convergence for Deep Neural Network Classifiers under the Teacher-Student Setting

Classifiers built with neural networks handle large-scale high-dimension...