from sagemaker import LinearLearner #import the LinearLearner model
#specify an output path for your model artifacts
prefix = 'fraud'
output_path = 's3://{}/{}'.format(bucket, prefix)
"""Instantiate LinearLearner model(set predictor type to binary classifier),
set positive example weight to balanced, AWS version of smote and set training target
recall to 90%(0.9)"""
linear_balanced = LinearLearner(role=role,
train_instance_count=1,
train_instance_type='ml.c4.xlarge',
predictor_type='binary_classifier',
output_path=output_path,
sagemaker_session=sagemaker_session,
epochs=15,
binary_classifier_model_selection_criteria='precision_at_target_recall', # target recall
target_recall=0.9, #tune the model for best recall value
positive_example_weight_mult='balanced')
"Convert train data to AWS algorithm data format -- record_set"
# convert features/labels to numpy
train_x_np = train_features.astype('float32')
train_y_np = train_labels.astype('float32')
# create RecordSet
formatted_train_data = linear.record_set(train_x_np, labels=train_y_np)
"train the estimator on formatted training data"
def test_linearlearner_airflow_config_uploads_data_source_to_s3(
sagemaker_session, cpu_instance_type
):
with timeout(seconds=AIRFLOW_CONFIG_TIMEOUT_IN_SECONDS):
data_path = os.path.join(DATA_DIR, "one_p_mnist", "mnist.pkl.gz")
pickle_args = {} if sys.version_info.major == 2 else {"encoding": "latin1"}
# Load the data into memory as numpy arrays
with gzip.open(data_path, "rb") as f:
train_set, _, _ = pickle.load(f, **pickle_args)
train_set[1][:100] = 1
train_set[1][100:200] = 0
train_set = train_set[0], train_set[1].astype(np.dtype("float32"))
ll = LinearLearner(
ROLE,
1,
cpu_instance_type,
predictor_type="binary_classifier",
sagemaker_session=sagemaker_session,
)
ll.binary_classifier_model_selection_criteria = "accuracy"
ll.target_recall = 0.5
ll.target_precision = 0.5
ll.positive_example_weight_mult = 0.1
ll.epochs = 1
ll.use_bias = True
ll.num_models = 1
ll.num_calibration_samples = 1
ll.init_method = "uniform"
ll.init_scale = 0.5
ll.init_sigma = 0.2
ll.init_bias = 5
ll.optimizer = "adam"
ll.loss = "logistic"
ll.wd = 0.5
ll.l1 = 0.5
ll.momentum = 0.5
ll.learning_rate = 0.1
ll.beta_1 = 0.1
ll.beta_2 = 0.1
ll.use_lr_scheduler = True
ll.lr_scheduler_step = 2
ll.lr_scheduler_factor = 0.5
ll.lr_scheduler_minimum_lr = 0.1
ll.normalize_data = False
ll.normalize_label = False
ll.unbias_data = True
ll.unbias_label = False
ll.num_point_for_scaler = 10000
ll.margin = 1.0
ll.quantile = 0.5
ll.loss_insensitivity = 0.1
ll.huber_delta = 0.1
ll.early_stopping_tolerance = 0.0001
ll.early_stopping_patience = 3
records = ll.record_set(train_set[0][:200], train_set[1][:200])
training_config = _build_airflow_workflow(
estimator=ll, instance_type=cpu_instance_type, inputs=records
)
_assert_that_s3_url_contains_data(
sagemaker_session,
training_config["InputDataConfig"][0]["DataSource"]["S3DataSource"]["S3Uri"],
)
# import LinearLearner
from sagemaker import LinearLearner
# instantiate LinearLearner
#Solution
# specify an output path
prefix = 'creditcard'
output_path = 's3://{}/{}'.format(bucket, prefix)
# instantiate LinearLearner
linear = LinearLearner(role=role,
train_instance_count=1,
train_instance_type='ml.c4.xlarge',
predictor_type='binary_classifier',
output_path=output_path,
sagemaker_session=sagemaker_session,
epochs=15)
# ### EXERCISE: Convert data into a RecordSet format
#
# Next, prepare the data for a built-in model by converting the train features and labels into numpy array's of float values. Then you can use the [record_set function](https://sagemaker.readthedocs.io/en/stable/linear_learner.html#sagemaker.LinearLearner.record_set) to format the data as a RecordSet and prepare it for training!
# In[13]:
# create RecordSet of training data
#formatted_train_data = None
# convert features/labels to numpy
train_x_np = train_features.astype('float32')
train_y_np = train_labels.astype('float32')