1. What is the recommended approach for optimizing the performance of AWS DeepLens during object detection tasks?

A) Increasing the batch size of input data to the neural network
B) Increasing the number of inference workers
C) Reducing the size of the input data to the neural network
D) Decreasing the number of inference workers



2. You have developed a machine learning model to predict customer churn for an e-commerce company, and you need to deploy the model to process new data in real-time. Which of the following AWS services would be the best fit for this use case?

A) Amazon SageMaker Batch Transform
B) Amazon SageMaker Endpoints
C) Amazon SageMaker Neo
D) AWS Lambda



3. A company wants to build a machine learning solution that can process data from various sources including IoT sensors, social media platforms, and databases. Which AWS services can be used to ingest and collect this data?

A) Amazon Kinesis Data Streams
B) Amazon Elastic File System (EFS)
C) Amazon Simple Storage Service (S3)
D) Amazon Simple Notification Service (SNS)
E) AWS Direct Connect


4. You are designing an Amazon Kinesis Data Analytics application to analyze a streaming data source from IoT devices. The IoT devices send data in JSON format, including sensor readings and device metadata. You want to extract specific fields from the JSON data and use them for aggregation and analysis. Which of the following is the most efficient approach to achieve this goal?

A) Use AWS Lambda to preprocess the JSON data and extract the required fields before ingesting it into Kinesis Data Analytics
B) Use the JSON SerDe to convert the JSON data into a tabular format, and then apply SQL transformations to extract the required fields
C) Use Kinesis Data Firehose to transform the JSON data into a tabular format, and then ingest it into Kinesis Data Analytics
D) Use the built-in JSON functions in Kinesis Data Analytics to extract the required fields directly from the JSON data



5. In Amazon Machine Learning, what is the purpose of hyperparameter tuning, and what techniques can be used to perform hyperparameter tuning?

A) Hyperparameter tuning is used to determine the optimal learning rate in a model, and it can be performed using grid search.
B) Hyperparameter tuning is used to determine the optimal regularization parameter in a model, and it can be performed using batch normalization.
C) Hyperparameter tuning is used to determine the optimal number of features in a model, and it can be performed using principal component analysis.
D) Hyperparameter tuning is used to determine the optimal number of iterations in a model, and it can be performed using randomized search.