1. Which of the following regularization techniques can be used to prevent overfitting in a machine learning model during training?

A) Dropout
B) Data augmentation
C) Gradient descent
D) Batch normalization
E) Early stopping


2. Which of the following is a valid way to configure an Amazon EKS cluster to use AWS Fargate?

A) Create a launch configuration with an IAM policy that allows access to the EKS control plane and the AWS Fargate profile, then launch an EC2 instance with that launch configuration.
B) Create an Amazon EKS cluster, create an AWS Fargate profile and associate it with the cluster, then deploy a Kubernetes pod to the Fargate profile.
C) Use the AWS CLI to create a task definition with a Fargate launch type, then create an Amazon EKS cluster and associate the task definition with the cluster.
D) Create a Kubernetes pod manifest that specifies the Fargate launch type, then deploy the manifest to an EC2 instance that is part of the EKS cluster.



3. Which of the following natural language processing (NLP) techniques is used for analyzing the sentiment of text data by assigning a polarity value to each word in the text?

A) Topic modeling
B) None of the above
C) Named entity recognition (NER)
D) Sentiment analysis
E) Part-of-speech tagging (POS tagging)


4. Which of the following evaluation metrics is best suited for measuring the performance of a natural language processing (NLP) model in sentiment analysis tasks?

A) Precision
B) None of the above
C) Recall
D) F1-score
E) Accuracy


5. Which of the following AWS services can be used to ensure the high availability and fault tolerance of a deployed machine learning model by automatically replicating data and workloads across multiple availability zones?

A) Amazon S3
B) AWS Glue
C) Amazon Redshift
D) Amazon Kinesis