AWS Certified Machine Learning Specialty Questions And Answers 2026/2027/2026 (Complete And Verified Study material) (2pages) LEARNEXAMS

What are Amazon Kinesis Video Streams? - Correct Answer- Amazon Kinesis Video Streams can be used to securely stream live video from devices to the AWS Cloud or build applications for real-time video processing and batch-oriented video analytics (ML, playback, and other processing). Kinesis Video Streams can also be used to watch our video streams in real time as they are received in the cloud. What is Amazon SageMaker Batch Transform? - Correct Answer- To preprocess or get inferences for an entire dataset, use Batch Transform. We use Batch Transform when we need to work with large datasets, process datasets quickly, or have subsecond latency. We use preprocessing to remove noise or bias from our dataset that interferes with training or inference. We use Batch Transform for inference when we don't need a persistent endpoint. For example, we can use Batch Transform to compare production variants that deploy different models. What is Keras? - Correct Answer- Keras is an open-source neural-network library written in Python. It is capable of running on top of TensorFlow, Microsoft Cognitive Toolkit, Theano, or PlaidML. Designed to enable fast experimentation with deep neural networks, it focuses on being user-friendly, modular, and extensible. What are the three stages of Amazon SageMaker? - Correct Answer- - Build (preprocessing, Amazon SageMaker Ground Truth, and notebooks) - Train (Using built-in and custom algorithms, hyperparameter tuning, notebooks, and infrastructure) - Deploy (Real-time and batch deployments, notebooks, infrastructure, and Amazon SageMaker Neo) What command is used to move data from an Amazon Redshift cluster to Amazon S3 using AWS glue? - Correct Answer- `UNLOAD` - Unloads the result of a query to one or more text files on Amazon S3. What is Gluon? - Correct Answer- Gluon is a clear, concise, and simple API for deep learning. It makes it easy to prototype, build, and train deep learning models without sacrificing training speed. What corrective actions can be taken when a model is underfitting the training data? - Correct Answer- Improve performance due to insufficient data by: - Increasing the amount of training data examples. - Increasing the number of passes on the existing training data. - Add new domain-specific features. - Add more Cartesian products. - Change the types eof feature processing (e.g. increase n-grams size). - Decrease the amount of regularization used. What are some use cases for Amazon Transcribe?