When Can Neural Networks Learn Connected Decision Regions?

Abstract

Previous work has questioned the conditions under which the decision regions of a neural network are connected and further showed the implications of the corresponding theory to the problem of adversarial manipulation of classifiers. It has been proven that for a class of activation functions including leaky ReLU, neural networks having a pyramidal structure, that is no layer has more hidden units than the input dimension, produce necessarily connected decision regions. In this paper, we advance this important result by further developing the sufficient and necessary conditions under which the decision regions of a neural network are connected. We then apply our framework to overcome the limits of existing work and further study the capacity to learn connected regions of neural networks for a much wider class of activation functions including those widely used, namely ReLU, sigmoid, tanh, softlus, and exponential linear function.

References in corpus (8)

• Auto-Encoding Variational Bayes
• The Loss Surfaces of Multilayer Networks
• The loss surface of deep and wide neural networks
• Beating the Perils of Non-Convexity: Guaranteed Training of Neural Networks using Tensor Methods
• Approximating Continuous Functions by ReLU Nets of Minimal Width
• Depth-Width Tradeoffs in Approximating Natural Functions with Neural Networks
• Optimization Landscape and Expressivity of Deep CNNs
• Theory II: Landscape of the Empirical Risk in Deep Learning