Steganografiya quyidagi sohalarda qo'llaniladi, lekin ular bilan cheklanmaydi

y,
test_size=0.1,
random_state=42)

# Train a watermarked model
print('Training watermarked model')
number_labels = len(np.unique([floor(k) for k in y_train]))
args = TrainingWMArgs(
nbr_classes=number_labels,
key_dawn=default_key(255),
probability_dawn=0.01,
trigger_technique='dawn',
metric='accuracy')

wm_model = Trainer(RandomForestClassifier(max_depth=1000, random_state=42),
args=args)
ownership = wm_model.fit(X_train, y_train)

verification = verify(
ownership['labels'],
wm_model.predict(ownership['inputs']),
number_labels=number_labels,
metric='accuracy',
dawn=True)

print(verification)


from math import floor
import numpy as np
from tqdm import tqdm

from sklearn.base import clone
from sklearn.model_selection import train_test_split
from mlmodelwatermarking.marklearn import Trainer
from mlmodelwatermarking import TrainingWMArgs
from mlmodelwatermarking.verification import verify

from warnings import simplefilter

simplefilter(action='ignore', category=FutureWarning)


def test_watermark_sklearn(X, y, base_model,
metric='accuracy', trigger_size=100):
""" Test the watermark functions

Parameters:
X (array): Input data
y (array): Label data (0 / 1)
base_model (Object): Model to be tested
metric (string): Type of metri for WM verification
trigger_size (int): Number of trigger inputs

"""
X_train, _, y_train, _ = train_test_split(X,
y,