apunkt/ai-privacy-toolkit
1
0
Fork 0
mirror of https://github.com/IBM/ai-privacy-toolkit.git synced 2026-07-05 16:02:10 +02:00
Code Issues Projects Releases Packages Wiki Activity Actions

New model wrappers (#32)

Browse source 
* keras wrapper + blackbox classifier wrapper (fix #7)

* fix error in NCP calculation

* Update notebooks

* Fix #25 (incorrect attack_feature indexes for social feature in notebook)

* Consistent naming of internal parameters
This commit is contained in:
abigailgold 2022-05-12 15:44:29 +03:00 • committed by GitHub
parent fd6be8e778
commit fe676fa426
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
15 changed files with 1407 additions and 656 deletions
Download patch file Download diff file Expand all files Collapse all files

54
tests/test_model.py
View file

@ -1,21 +1,24 @@
import pytest
from apt.utils.models import SklearnClassifier, SklearnRegressor, ModelOutputType
from apt.utils.datasets import ArrayDataset
from apt.utils.models import SklearnClassifier, SklearnRegressor, ModelOutputType, KerasClassifier, BlackboxClassifier
from apt.utils.datasets import ArrayDataset, Data
from apt.utils import dataset_utils
from sklearn.tree import DecisionTreeRegressor
from sklearn.ensemble import RandomForestClassifier
from tensorflow.keras.models import Sequential
from tensorflow.keras.layers import Dense, Input
def test_sklearn_classifier():
(x_train, y_train), (x_test, y_test) = dataset_utils.get_iris_dataset()
(x_train, y_train), (x_test, y_test) = dataset_utils.get_iris_dataset_np()
underlying_model = RandomForestClassifier()
model = SklearnClassifier(underlying_model, ModelOutputType.CLASSIFIER_VECTOR)
model = SklearnClassifier(underlying_model, ModelOutputType.CLASSIFIER_PROBABILITIES)
train = ArrayDataset(x_train, y_train)
test = ArrayDataset(x_test, y_test)
model.fit(train)
pred = model.predict(x_test)
pred = model.predict(test)
assert(pred.shape[0] == x_test.shape[0])
score = model.score(test)
@ -23,13 +26,50 @@ def test_sklearn_classifier():
def test_sklearn_regressor():
(x_train, y_train), (x_test, y_test) = dataset_utils.get_diabetes_dataset()
(x_train, y_train), (x_test, y_test) = dataset_utils.get_diabetes_dataset_np()
underlying_model = DecisionTreeRegressor()
model = SklearnRegressor(underlying_model)
train = ArrayDataset(x_train, y_train)
test = ArrayDataset(x_test, y_test)
model.fit(train)
pred = model.predict(x_test)
pred = model.predict(test)
assert (pred.shape[0] == x_test.shape[0])
score = model.score(test)
def test_keras_classifier():
(x_train, y_train), (x_test, y_test) = dataset_utils.get_iris_dataset_np()
underlying_model = Sequential()
underlying_model.add(Input(shape=(4,)))
underlying_model.add(Dense(100, activation="relu"))
underlying_model.add(Dense(10, activation="relu"))
underlying_model.add(Dense(3, activation='softmax'))
underlying_model.compile(loss="categorical_crossentropy", optimizer="adam", metrics=["accuracy"])
model = KerasClassifier(underlying_model, ModelOutputType.CLASSIFIER_PROBABILITIES)
train = ArrayDataset(x_train, y_train)
test = ArrayDataset(x_test, y_test)
model.fit(train)
pred = model.predict(test)
assert(pred.shape[0] == x_test.shape[0])
score = model.score(test)
assert(0.0 <= score <= 1.0)
def test_blackbox_classifier():
(x_train, y_train), (x_test, y_test) = dataset_utils.get_iris_dataset_np()
train = ArrayDataset(x_train, y_train)
test = ArrayDataset(x_test, y_test)
data = Data(train, test)
model = BlackboxClassifier(data, ModelOutputType.CLASSIFIER_PROBABILITIES)
pred = model.predict(test)
assert(pred.shape[0] == x_test.shape[0])
score = model.score(test)
assert(0.0 <= score <= 1.0)