NeuroSCA: Evolving Activation Functions for Side-Channel Analysis

Abstract

The choice of activation functions can significantly impact the performance of neural networks. Due to an ever-increasing number of new activation functions being proposed in the literature, selecting the appropriate activation function becomes even more difficult. Consequently, many researchers approach this problem from a different angle, in which instead of selecting an existing activation function, an appropriate activation function is evolved for the problem at hand. In this paper, we demonstrate that evolutionary algorithms can evolve new activation functions for side-channel analysis (SCA), outperforming ReLU and other activation functions commonly applied to that problem. More specifically, we use Genetic Programming to define and explore candidate activation functions (neuroevolution) in the form of mathematical expressions that are gradually improved. Experiments with the ASCAD database show that this approach is highly effective compared to results obtained with standard activation functions and that it can match the stateof-the-art results from the literature. More precisely, the obtained results for the ASCAD fixed key dataset demonstrate that the evolved activation functions can improve the current state-of-the-art by achieving a guessing entropy of 287 for the Hamming weight model and 115 for the Identity leakage model, compared to 447 and 120 obtained in the literature.

Texture-Based Classification of Benign and Malignant Mammography Images using Weka Machine Learning: An Optimal Approach

“…One of the main drawbacks of the Naive Bayes classification is its vital independence feature. In practice, having set of entirely independent features is almost impossible [33][34][35][36] . In classifying benign and malignant image data on mammography, the texture features are not entirely independent and even overlap.…”

“…Feature scaling is a method of having numerical data in a dataset with the same values range. MLP is proven to be able to separate datasets well with up to 100% accuracy [35][36][37] . The results of texture feature extraction on benign and malignant images in this study show a range that is not too far away so that some similar data are grouped into the same scale.…”

NASCTY: Neuroevolution to Attack Side-channel Leakages Yielding Convolutional Neural Networks

“…The application of neuroevolution to perform side-channel analysis has only scarcely been explored in existing work. Knezevic et al used genetic programming to evolve custom activation functions specific to side-channel analysis [9] that can outperform the widely used ReLU function. The genome in their approach encodes an activation function as a tree containing unary and binary operators, with leaves representing the function's inputs.…”

An analysis method for improvement of coefficient consumption of related coefficients

“…This article only discusses non model-building attack. Common non model-building attack methods include simple power consumption attack (SPA), differential power consumption analysis attack (DPA), related power consumption analysis attack (CPA),etc [4] .…”