verify.ho: verify.ho

In sideChannelAttack: Side Channel Attack

Description

The verify.ho function allows to estimate the quality of a model (an attack) with a hold-out approach.

Usage

verify.ho(model, filter, Xlearn, Ylearn, Xval, Yval, nbreVarX, param.model=list(), param.fs=list(), ...)

Arguments

model	A model such as randomForest, gaussian, svm, etc.
filter	A feature selection such as filter.PCA, filter.mRMR, etc.
Xlearn	A matrix where each row is a physical measure for the training set.
Ylearn	A vector where the i_th element of the vector Ylearn is the key for the i_th physical measure in the matrix Xlearn.
Xval	A matrix where each row is a physical measure for the validation set.
Yval	A vector where the i_th element of the vector Yval is the key for the i_th physical measure in the matrix Xval.
nbreVarX	The number of variables which represents each physical measure after the reduction by the feature selection.
param.model	A list of parameters to insert into the model.
param.fs	A list of parameters to insert into the feature selection algorithm.
...	Currently ignored.

Details

The verify.ho function estimates the quality of the attack with a hold-out approach. It cuts randomly the set of observations in two subsets. The first one is retained as the training and the second is used for the validation phase where the quality of the model is estimatedby counting the number of time where model predicts correctly the key.

Value

The verify.ho function returns an object which can be used with the plot function to plot the quality of the model.

The value of this function is an object of class verify.ho, which is a list with the following components:

TP	number of true positive
TN	number of true negative
FN	number of false negative
FP	number of false positive
dim	the number of variables which represents each physical measure after the reduction by the feature selection.

Author(s)

Liran Lerman [e-mail] & Gianluca Bontempi [e-mail]@ulb.ac.be & Olivier Markowitch [e-mail]

Examples

#data collection
data(powerC)
traces = powerC[,-301]
traces = powerC[,1:10]
key = powerC[,301]
newIndice = sample(1:256)
traces = traces[newIndice,]
key = key[newIndice]+1

#model checking
attack=verify.ho(model=gaussian,filter=filter.PCA,Xlearn=traces[1:128,],Ylearn=key[1:128],Xval=traces[128:256,],Yval=key[128:256],nbreVarX=c(3:4))
plot(attack)

sideChannelAttack documentation built on May 30, 2017, 7:24 a.m.

verify.cv

sideChannelAttack index

verify.loo