Confuse Skull: SageMaker - Amazon ML MCQ's

SageMaker - Amazon ML MCQ's

1. Asynchronous Predictions are possible in SageMaker through _________________.

BatchTranform

2. Training Data for SageMaker models is ___________________.

Stored in S3

3. Which of the following algorithms is best suited for detecting malicious server activity?

IP

4. Hyperparameters are _______________________.

Model-specific parameters which are preset before training

5. Hold-Out dataset is used for _________________.

Validation

6. Inference code _________________________.

Gets predictions from a model

7. SageMaker metrics are captured by ______________.

CloudWatch

8. Custom Algorithms can be implemeted via _________________.

Docker and Jupyter

9. Hyperparameter tuning in SageMaker is ___________________.

Automatic once enabled

10. Which of the following is a statically typed language preferred for performance in Machine Learning tasks?

Scala

11. Metrics are used to ____________________.

Control Auto-scaling and manage search results

12. Automated Hyperparameter tuning is not recommended for ___________________.

all

13. It is mandatory to capture metrics for custom models ________________.

To specify logged data and use regular expressions

14. Log scaled parameters for hyperparameter tuning _____________________.

Can be specified or automatically detected

15. Auto-Scaling is suited for ____________________.

Production Environment

16. We can ensure models are secure, at ease, _________________________________.

when they are pre-approved in MarketPlace

17. Pause and Resume of Hyperparameter tuning jobs are _______________.

Supported by a limited number of algorithms

18. Cross-Platform Model Deployment for inferences is enabled through ______________.

SageMaker Neo

19. Search results of models can be traced to ___________________.

ALL

20. Automated Hyperparameter tuning is not recommended for ___________________.

All the given options

21. Reinforcement Learning optimizes agent performance in an environment with help of__________________.

Reward and punishment

22. Amazon Cognito helps in private workforce management through ___________.

Teams and Tasks

23. Most of the SageMaker resources like instances and communication between resources are internet-enabled by default.

True

24. Greengrass is used to deploy models for _________________.

IOT

25. Multiple label classification can be facilitated using Ground Truth.

Annotation Consolidation

26. Which of the following statements is not correct with respect to SageMakerFullAccess Policy?

It gives access to all related S3 buckets.

27. Ground Truth Labeling is ___________________.

ALl

28. API calls for instance management is recorded by ________________.

CloudTrail

29. MDPs state helps in _________________.

All

30. Artifacts from training jobs in S3 can be deleted once the model is deployed saving space and money.

True

31. Auto Scaling of SageMaker Instances is controlled by ________________.

CloudWatch

32. GPU Acceleration improves the performance of Instances.

Depends on Instance size

33. In GroundTruth, to reduce costs, manually labeled data can be used for ______________.

All

34. Instance Access in SageMaker can be restricted using __________________.

IAM

35. Metrics can be searched actively before _____________ of their latest update.

2 weeks

36. Which of the following service can be used for Data Collection and Transformation?

S3

37. Default instance storage allotted is _______________________.

50 GB

38. Overfitting can be avoided by increasing ________________________.

Hyperparameter Regulation

39. Identify the Memory-based Learning models.

K

40. Containers in Inference Pipelines communicate with each other using ________________.

SSL connection

3 comments:

UnknownNovember 20, 2021 at 9:24 PM
Thank you for the answer....!
ReplyDelete
Replies
UnknownMarch 10, 2022 at 11:30 AM
Hi bro thanks and some answers are wrong
32 wrong
30false
ReplyDelete
Replies
Mrinmoyee BhattacharyaMarch 14, 2022 at 3:34 PM
Reinforcement Learning models can be trained in simulated VR environments using
ReplyDelete
Replies