Analysis of Generalizability of Deep Neural Networks Based on the Complexity of Decision Boundary

Abstract

For supervised learning models, the analysis of generalization ability (generalizability) is vital because the generalizability expresses how well a model will perform on unseen data. Traditional generalization methods, such as the VC dimension, do not apply to deep neural network (DNN) models. Thus, new theories to explain the generalizability of DNNs are required. In this study, we hypothesize that the DNN with a simpler decision boundary has better generalizability by the law of parsimony (Occam's Razor). We create the decision boundary complexity (DBC) score to define and measure the complexity of decision boundary of DNNs. The idea of the DBC score is to generate data points (called adversarial examples) on or near the decision boundary. Our new approach then measures the complexity of the boundary using the entropy of eigenvalues of these data. The method works equally well for high-dimensional data. We use training data and the trained model to compute the DBC score. And, the ground truth for model's generalizability is its test accuracy. Experiments based on the DBC score have verified our hypothesis. The DBC is shown to provide an effective method to measure the complexity of a decision boundary and gives a quantitative measure of the generalizability of DNNs.

arXiv comment7 pages, 11 figures. Accepted by ICMLA 2020

References in corpus (10)

• Exploring Generalization in Deep Learning
• Computing Nonvacuous Generalization Bounds for Deep (Stochastic) Neural Networks with Many More Parameters than Training Data
• Spectral Norm Regularization for Improving the Generalizability of Deep Learning
• Fantastic Generalization Measures and Where to Find Them
• Predicting the Generalization Gap in Deep Networks with Margin Distributions
• Characterizing the Decision Boundary of Deep Neural Networks
• On the Decision Boundary of Deep Neural Networks
• The intriguing role of module criticality in the generalization of deep networks
• Understanding Why Neural Networks Generalize Well Through GSNR of Parameters
• Investigating Decision Boundaries of Trained Neural Networks

Cited by in corpus (2)

• Understanding Deep Learning via Decision Boundary
• The Missing Margin: How Sample Corruption Affects Distance to the Boundary in ANNs