1. A company is building a sentiment analysis model using Amazon Comprehend to analyze customer reviews of their products. The company wants to train the model using custom labels to better categorize customer sentiment. Which of the following is the correct process for creating custom labels in Amazon Comprehend?

A) Use the AWS SDK or AWS CLI to create a custom classification endpoint, and then use this endpoint to label the data.
B) Use the Amazon Comprehend console to create a custom classification job, and then use this job to label the data.
C) Use Amazon S3 to upload a file with the custom labels, and then use this file to label the data in Amazon Comprehend.
D) Use Amazon Translate to translate the data into custom labels, and then use these labels to train the sentiment analysis model.



2. Which of the following statements is true about the AWS Deep Learning AMIs (DLAMI)?

A) The AWS Deep Learning AMIs include popular deep learning frameworks, such as TensorFlow, PyTorch, and MXNet, pre-installed and optimized for use on Amazon EC2 instances.
B) AWS Deep Learning AMIs can only be used on instances with GPUs, and cannot be used on instances with CPUs.
C) AWS Deep Learning AMIs are only available for use in the us-east-1 and us-west-2 regions.
D) AWS Deep Learning AMIs require customers to have deep learning expertise and cannot be used by users without prior experience.



3. Which of the following statements is true about AWS DeepLens?

A) AWS DeepLens is a deep learning-enabled video camera for developers.
B) AWS DeepLens is a tool used to train models using reinforcement learning.
C) AWS DeepLens is a virtual reality headset developed by AWS.
D) AWS DeepLens is a serverless analytics service.
E) AWS DeepLens is a hardware appliance for managing Amazon S3.


4. In Amazon Forecast, what is the difference between a predictor and a forecast?

A) A predictor is an algorithm that learns from historical data to make predictions, while a forecast is the actual prediction generated by the predictor.
B) A predictor is a visual representation of historical data, while a forecast is the prediction generated by the predictor.
C) A predictor is a feature that enables you to group data by shared attributes, while a forecast is the final output generated by Amazon Forecast.
D) A predictor is a model that you train to make predictions based on historical data, while a forecast is a type of visualization of the data.



5. Which of the following is a correct statement about using Amazon Fraud Detector for detecting fraud in your business?

A) Amazon Fraud Detector integrates with AWS services such as Amazon SageMaker, Amazon SNS, and Amazon CloudWatch.
B) Amazon Fraud Detector supports only supervised learning techniques for detecting fraud.
C) Amazon Fraud Detector does not require any data preparation or feature engineering for detecting fraud.
D) Amazon Fraud Detector can detect fraud in any industry except for the finance industry.
E) Amazon Fraud Detector is not scalable for large datasets.