0% Complete
Home
/
13th International Conference on Computer and Knowledge Engineering
A large input-space-margin approach for adversarial training
Authors :
Reihaneh Nikouei
1
Mohammad Taheri
2
1- Department of Computer Science and Engineering, Shiraz University, Shiraz, Iran.
2- Department of Computer Science and Engineering, Shiraz University, Shiraz, Iran.
Keywords :
Adversarial attack،Large margin،Input space،Defense method
Abstract :
It is shown that machine learning models are vulnerable to adversarial attacks. Therefore, different defense methods such as adversarial training have been proposed to improve models’ robustness against these attacks. Some recent approaches proposed to structurally improve the robustness of the models. For example, large margin methods try to increase a margin, empty of instances, along decision boundaries that structurally increase necessary change to modify a training instance to an adversarial one. However, nonlinear large-margin models, maximize the margin in a high dimensional space although adversarial examples are generated with a little change in the original space. In this paper, a novel mixed approach is proposed, called LIM (Large Input Margin) to improve the robustness of the model by minimizing both structural and empirical risks. Specifically, both training and adversarial example generation are done based on a loss function to maximize the margin in the original feature space even in a non-linear model. The proposed method is evaluated with FGSM and PGD attacks on MNIST and CIFAR10 datasets. The experimental results show that LIM method outperforms the state-of-the-art defense methods significantly and improves adversarial robustness against FGSM and PGD attacks on both datasets.
Papers List
List of archived papers
Artificial Intelligence applications addressing different aspects of the Covid-19 crisis and key technological solutions for future epidemics control
Nadia Khalili - Hojatollah Hamidi
Improvement of Credit Scoring by LSTM Autoencoder Model
Milad Sattari Maleki - Seyedeh Niusha Motevallian - Faezehsadat Hosseini - Mohammad Sabokrou - Hamidreza Soltanalizadeh Maleki
Crack Segmentation in Civil Structure Images Using a Deep Learning Based Multi-Classifier System
Mohammadreza Asadi - Seyedeh Sogand Hashemi - Mohammad Taghi Sadeghi
A Review on Machine Learning Methods for Workload Prediction in Cloud Computing
Mohammad Yekta - Hadi Shahriar Shahhoseini
Divide and Conquer Approach to Long Genomic Sequence Alignment
Mahmoud Naghibzadeh - Samira Babaei - Behshid Behkmal - Mojtaba Hatami
The process of multi class fake news dataset generation
Sajjad Rezaei - Mohsen Kahani - Behshid Behkamal
Analysis of Insect-plant Interactions Affected by Mining operations, A Graph Mining Approach
Mohammad Heydari - Ali Bayat - Amir Albadvi
Early detection of Parkinson’s disease using Convolutional Neural Networks on SPECT images
Reyhaneh Dehghan - Marjan Naderan - Seyyed Enayatallah Alavi
Hybrid navigation based on GPS data and SIFT-based place recognition using Biologically-inspired SLAM
Sahar Salimpour Kasebi - Hadi Seyedarabi - Javad Musevi Niya
PowerLinear Activation Functions with application to the first layer of CNNs
Kamyar Nasiri - Kamaledin Ghiasi-Shirazi
more
Samin Hamayesh - Version 41.3.1