Attend the Women in ML Symposium on December 7 Register now

tf_privacy.synthetic_linearly_separable_data

Stay organized with collections Save and categorize content based on your preferences.

Generates synthetic train and test data for logistic regression.

num_train number of training data points.
num_test number of test data points.
dimension the dimension of the classification problem.
num_classes number of classes, assumed to be at least 2.

train_dataset num_train labeled examples, with unit l2-norm points.
test_dataset num_test labeled examples, with unit l2-norm points.