Adaptive Retraining for Neural Network Robustness in Classification
| Autor: | Ke Pei, Ruozhu Yao, Zheng Hu, Chengqiang Huang |
|---|---|
| Rok vydání: | 2021 |
| Předmět: |
Artificial neural network
business.industry Computer science Retraining Machine learning computer.software_genre Task (project management) Adversarial system ComputingMethodologies_PATTERNRECOGNITION Robustness (computer science) Decision boundary Artificial intelligence Applications of artificial intelligence business Focus (optics) computer |
| Zdroj: | IJCNN |
| DOI: | 10.1109/ijcnn52387.2021.9534294 |
| Popis: | Neural network robustness has been a hot research topic since the discovery of adversarial examples. With the increasing threats posed by adversarial examples in safety-critical AI applications, the research problem of how to boost the robustness of neural networks has drawn even more attention in recent years. In this paper, we argue that existing methods for neural network robustness optimization primarily focus on increasing the minimum adversarial perturbation of individual datum while neglecting the purpose of the target machine learning task, i.e., classification. Consequently, we propose a neural network retraining method which implicitly enhances a model's capability in maximizing the minimum distance from data instances of all classes to the decision boundary. As demonstrated by the experiments, our method succeeds in improving the robustness of pre-trained neural networks within only a few retraining epochs while imposing little negative effect on classification accuracy, thus being applicable for online optimization of neural networks. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|