code for membershipinference against machine learning paper:https://ieeexplore.ieee.org/abstract/document/7958568
use pytorch(GPU version) data:CIFAR-10
85%.py:nn model CIFAR10.py:nn model
traintarget.py:train target model
targetmodel_test.py:get accuracy of the target model and generate data for attack model test
cifartoimg.py:convert cifar data(python version) to img data
dataloader.py:dataloader for target/shadow model
selectdata_shadow.py:select data for shadow model train/test
trainshadow.py:trian shadow model on the data that from selectdata_shadow.py
shadowmodel_test.py:get accuracy of the trained shadow model and generate train data for attack model
trainattack.py:train attack model on the data that from shadowmodel_test.py
attackmodel_test.py:get accuracy of the attack model