scholarly journals Improving CEMA using Correlation Optimization

2018 ◽  
pp. 1-24 ◽  
Author(s):  
Pieter Robyns ◽  
Peter Quax ◽  
Wim Lamotte
Keyword(s):  
Machine Learning ◽  
Loss Function ◽  
Time Domain ◽  
Optimization Problem ◽  
Pearson Correlation ◽  
High Dimensional ◽  
Electromagnetic Analysis ◽  
Novel Approach ◽  
Secret Keys ◽  
Selection Of

Sensitive cryptographic information, e.g. AES secret keys, can be extracted from the electromagnetic (EM) leakages unintentionally emitted by a device using techniques such as Correlation Electromagnetic Analysis (CEMA). In this paper, we introduce Correlation Optimization (CO), a novel approach that improves CEMA attacks by formulating the selection of useful EM leakage samples in a trace as a machine learning optimization problem. To this end, we propose the correlation loss function, which aims to maximize the Pearson correlation between a set of EM traces and the true AES key during training. We show that CO works with high-dimensional and noisy traces, regardless of time-domain trace alignment and without requiring prior knowledge of the power consumption characteristics of the cryptographic hardware. We evaluate our approach using the ASCAD benchmark dataset and a custom dataset of EM leakages from an Arduino Duemilanove, captured with a USRP B200 SDR. Our results indicate that the masked AES implementation used in all three ASCAD datasets can be broken with a shallow Multilayer Perceptron model, whilst requiring only 1,000 test traces on average. A similar methodology was employed to break the unprotected AES implementation from our custom dataset, using 22,000 unaligned and unfiltered test traces.

scholarly journals A Multicriteria Approach to Find Predictive and Sparse Models with Stable Feature Selection for High-Dimensional Data

2017 ◽  
Vol 2017 ◽  
pp. 1-18 ◽  
Author(s):  
Andrea Bommert ◽  
Jörg Rahnenführer ◽  
Michel Lang
Keyword(s):  
Feature Selection ◽  
Predictive Model ◽  
Predictive Accuracy ◽  
Pearson Correlation ◽  
High Dimensional Data ◽  
High Dimensional ◽  
Sparse Models ◽  
Data Set ◽  
The Stability ◽  
Selection Of

Finding a good predictive model for a high-dimensional data set can be challenging. For genetic data, it is not only important to find a model with high predictive accuracy, but it is also important that this model uses only few features and that the selection of these features is stable. This is because, in bioinformatics, the models are used not only for prediction but also for drawing biological conclusions which makes the interpretability and reliability of the model crucial. We suggest using three target criteria when fitting a predictive model to a high-dimensional data set: the classification accuracy, the stability of the feature selection, and the number of chosen features. As it is unclear which measure is best for evaluating the stability, we first compare a variety of stability measures. We conclude that the Pearson correlation has the best theoretical and empirical properties. Also, we find that for the stability assessment behaviour it is most important that a measure contains a correction for chance or large numbers of chosen features. Then, we analyse Pareto fronts and conclude that it is possible to find models with a stable selection of few features without losing much predictive accuracy.

Generation of Periodic Trajectories for Optimal Robot Excitation

1997 ◽  
Vol 119 (4A) ◽  
pp. 611-615 ◽  
Author(s):  
J. Swevers ◽  
C. Ganseman ◽  
J. De Schutter ◽  
H. Van Brussel
Keyword(s):  
Fourier Series ◽  
Time Domain ◽  
Optimization Problem ◽  
Identification Accuracy ◽  
Condition Numbers ◽  
Frequency Content ◽  
Motion Constraints ◽  
Cartesian Space ◽  
Time Domain Data ◽  
Selection Of

This paper describes the parameterization of robot excitation trajectories for optimal robot identification based on finite Fourier series. The coefficients of the Fourier series are optimized for minimal sensitivity of the identification to measurement disturbances, which is measured as the condition number of a regression matrix, taking into account motion constraints in joint and cartesian space. This approach allows obtaining small condition numbers with few coefficients for each joint, which simplifies the optimization problem significantly. The periodicity of the resulting trajectories and the fact that one has total control over their frequency content, are additional features of the presented parameterization approach. They allow further optimization of the excitation experiments through time domain data averaging and optimal selection of the excitation bandwidth, which both help the reduction of the disturbance level on the measurements, and therefore improve the identification accuracy.

scholarly journals Sparse Data–Driven Learning for Effective and Efficient Biomedical Image Segmentation

2020 ◽  
Vol 22 (1) ◽  
pp. 127-153
Author(s):  
John A. Onofrey ◽  
Lawrence H. Staib ◽  
Xiaojie Huang ◽  
Fan Zhang ◽  
Xenophon Papademetris ◽  
...  
Keyword(s):  
Machine Learning ◽  
Image Segmentation ◽  
Deep Learning ◽  
Computational Efficiency ◽  
Medical Image ◽  
Medical Image Segmentation ◽  
Data Driven ◽  
High Dimensional ◽  
Biomedical Image ◽  
Selection Of

Sparsity is a powerful concept to exploit for high-dimensional machine learning and associated representational and computational efficiency. Sparsity is well suited for medical image segmentation. We present a selection of techniques that incorporate sparsity, including strategies based on dictionary learning and deep learning, that are aimed at medical image segmentation and related quantification.

scholarly journals Analysis and object markup of hyperspectral images for machine learning methods

2019 ◽  
pp. 309-317
Author(s):  
V P Gromov ◽  
L I Lebedev ◽  
V E Turlapov
Keyword(s):  
Machine Learning ◽  
Pearson Correlation ◽  
Principal Component ◽  
Hyperspectral Images ◽  
Spectral Features ◽  
Two Dimensional ◽  
Learning Methods ◽  
Machine Learning Methods ◽  
Reference Signature ◽  
Selection Of

The development of the nominal sequence of steps for analyzing the HSI proposed by Landgrebe, which is necessary in the context of the appearance of reference signature libraries for environmental monitoring, is discussed. The approach is based on considering the HSI pixel as a signature that stores all spectral features of an object and its states, and the HSI as a whole - as a two-dimensional signature field. As a first step of the analysis, a procedure is proposed for detecting a linear dependence of signatures by the magnitude of the Pearson correlation coefficient. The main apparatus of analysis, as in Landgrebe sequence, is the method of principal component analysis, but it is no longer used to build classes and is applied to investigate the presence in the class of subclasses essential for the applied area. The experimental material includes such objects as water, swamps, soil, vegetation, concrete, pollution. Selection of object samples on the image is made by the user. From the studied images of HSI objects, a base of reference signatures for classes (subclasses) of objects is formed, which in turn can be used to automate HSI markup with the aim of applying machine learning methods to recognize HSI objects and their states.

scholarly journals Pitting Judgment Model Based on Machine Learning and Feature Optimization Methods

2021 ◽  
Vol 8 ◽  
Author(s):  
Zhihao Qu ◽  
Dezhi Tang ◽  
Zhu Wang ◽  
Xiaqiao Li ◽  
Hongjian Chen ◽  
...  
Keyword(s):  
Machine Learning ◽  
Pipeline Steel ◽  
Pearson Correlation ◽  
Feature Reduction ◽  
Oil Field ◽  
High Dimensional ◽  
Good Prediction ◽  
Model Based ◽  
Good Prediction Accuracy ◽  
Corrosion Data

Pitting corrosion seriously harms the service life of oil field gathering and transportation pipelines, which is an important subject of corrosion prevention. In this study, we collected the corrosion data of pipeline steel immersion experiment and established a pitting judgment model based on machine learning algorithm. Feature reduction methods, including feature importance calculation and pearson correlation analysis, were first adopted to find the important factors affecting pitting. Then, the best input feature set for pitting judgment was constructed by combining feature combination and feature creation. Through receiver operating characteristic (ROC) curve and area under curve (AUC) calculation, random forest algorithm was selected as the modeling algorithm. As a result, the pitting judgment model based on machine learning and high dimensional feature parameters (i.e., material factors, solution factors, environment factors) showed good prediction accuracy. This study provided an effective means for processing high-dimensional and complex corrosion data, and proved the feasibility of machine learning in solving material corrosion problems.

Fast Time Domain Integral Equation Solvers for Large-Scale Electromagnetic Analysis

2004 ◽  
Author(s):  
Eric Michielsssen ◽  
Weng C. Chew ◽  
Jianming Jin ◽  
Balasubramaniam Shanker
Keyword(s):  
Integral Equation ◽  
Time Domain ◽  
Large Scale ◽  
Electromagnetic Analysis ◽  
Fast Time ◽  
Domain Integral ◽  
Time Domain Integral Equation

Deep Learning in Disease Diagnosis: Models and Datasets

2020 ◽  
Vol 15 ◽  
Author(s):  
Deeksha Saxena ◽  
Mohammed Haris Siddiqui ◽  
Rajnish Kumar
Keyword(s):  
Biological Sciences ◽  
Machine Learning ◽  
Deep Learning ◽  
Disease Diagnosis ◽  
Learning Models ◽  
Data Types ◽  
Related Data ◽  
Abstract Level ◽  
Experimental Validations ◽  
Selection Of

Background: Deep learning (DL) is an Artificial neural network-driven framework with multiple levels of representation for which non-linear modules combined in such a way that the levels of representation can be enhanced from lower to a much abstract level. Though DL is used widely in almost every field, it has largely brought a breakthrough in biological sciences as it is used in disease diagnosis and clinical trials. DL can be clubbed with machine learning, but at times both are used individually as well. DL seems to be a better platform than machine learning as the former does not require an intermediate feature extraction and works well with larger datasets. DL is one of the most discussed fields among the scientists and researchers these days for diagnosing and solving various biological problems. However, deep learning models need some improvisation and experimental validations to be more productive. Objective: To review the available DL models and datasets that are used in disease diagnosis. Methods: Available DL models and their applications in disease diagnosis were reviewed discussed and tabulated. Types of datasets and some of the popular disease related data sources for DL were highlighted. Results: We have analyzed the frequently used DL methods, data types and discussed some of the recent deep learning models used for solving different biological problems. Conclusion: The review presents useful insights about DL methods, data types, selection of DL models for the disease diagnosis.

scholarly journals Classification of Brainwaves for Sleep Stages by High-Dimensional FFT Features from EEG Signals

2020 ◽  
Vol 10 (5) ◽  
pp. 1797 ◽  
Author(s):  
Mera Kartika Delimayanti ◽  
Bedy Purnama ◽  
Ngoc Giang Nguyen ◽  
Mohammad Reza Faisal ◽  
Kunti Robiatul Mahmudah ◽  
...  
Keyword(s):  
Machine Learning ◽  
Sleep Stage ◽  
Machine Learning Algorithms ◽  
High Dimensional ◽  
Sleep Stages ◽  
Eeg Signals ◽  
Stage Classification ◽  
Sleep Stage Classification ◽  
Low Dimensional

Manual classification of sleep stage is a time-consuming but necessary step in the diagnosis and treatment of sleep disorders, and its automation has been an area of active study. The previous works have shown that low dimensional fast Fourier transform (FFT) features and many machine learning algorithms have been applied. In this paper, we demonstrate utilization of features extracted from EEG signals via FFT to improve the performance of automated sleep stage classification through machine learning methods. Unlike previous works using FFT, we incorporated thousands of FFT features in order to classify the sleep stages into 2–6 classes. Using the expanded version of Sleep-EDF dataset with 61 recordings, our method outperformed other state-of-the art methods. This result indicates that high dimensional FFT features in combination with a simple feature selection is effective for the improvement of automated sleep stage classification.

Sign in / Sign up

Export Citation Format

Share Document