TY - GEN
T1 - Learning When to Stop
T2 - 12th International Workshop on Constructive Side-Channel Analysis and Secure Design, COSADE 2021
AU - Perin, Guilherme
AU - Buhan, Ileana
AU - Picek, Stjepan
PY - 2021
Y1 - 2021
N2 - Today, deep neural networks are a common choice for conducting the profiled side-channel analysis. Unfortunately, it is not trivial to find neural network hyperparameters that would result in top-performing attacks. The hyperparameter leading the training process is the number of epochs during which the training happens. If the training is too short, the network does not reach its full capacity, while if the training is too long, the network overfits and cannot generalize to unseen examples. In this paper, we tackle the problem of determining the correct epoch to stop the training in the deep learning-based side-channel analysis. We demonstrate that the amount of information, or, more precisely, mutual information transferred to the output layer, can be measured and used as a reference metric to determine the epoch at which the network offers optimal generalization. To validate the proposed methodology, we provide extensive experimental results.
AB - Today, deep neural networks are a common choice for conducting the profiled side-channel analysis. Unfortunately, it is not trivial to find neural network hyperparameters that would result in top-performing attacks. The hyperparameter leading the training process is the number of epochs during which the training happens. If the training is too short, the network does not reach its full capacity, while if the training is too long, the network overfits and cannot generalize to unseen examples. In this paper, we tackle the problem of determining the correct epoch to stop the training in the deep learning-based side-channel analysis. We demonstrate that the amount of information, or, more precisely, mutual information transferred to the output layer, can be measured and used as a reference metric to determine the epoch at which the network offers optimal generalization. To validate the proposed methodology, we provide extensive experimental results.
KW - Information bottleneck
KW - Mutual information
KW - Neural networks
KW - Overfitting
KW - Side-channel analysis
UR - http://www.scopus.com/inward/record.url?scp=85118966979&partnerID=8YFLogxK
U2 - 10.1007/978-3-030-89915-8_3
DO - 10.1007/978-3-030-89915-8_3
M3 - Conference contribution
AN - SCOPUS:85118966979
SN - 9783030899141
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 53
EP - 81
BT - Constructive Side-Channel Analysis and Secure Design - 12th International Workshop, COSADE 2021, Proceedings
A2 - Bhasin, Shivam
A2 - De Santis, Fabrizio
PB - Springer Science+Business Media
Y2 - 25 October 2021 through 27 October 2021
ER -