Model Training and Tuning Flashcards

AWS Machine Learning Engineer Associate MLA-C01 Flashcards

Front	Back
How can SageMaker manage custom training scripts	By packaging the scripts in a Docker container and running it on SageMaker.
How does SageMaker Clarify assist with fairness	By detecting biases in data and explaining predictions of machine learning models.
How does SageMaker Data Wrangler assist with data preparation	By providing tools to import, clean, transform, and visualize data for machine learning.
How does SageMaker handle multi-model endpoints	By hosting multiple models under a single endpoint and dynamically loading them as needed.
How does SageMaker help with data annotation	Through its built-in labeling services called SageMaker Ground Truth.
How does SageMaker simplify distributed training	By automatically managing the infrastructure for scaling training across multiple instances.
How is active learning used in SageMaker Ground Truth	A method where the model selects the most uncertain data points for labeling to improve efficiency.
What are data augmentations in model training	Techniques to improve model performance by augmenting the training data.
What does SageMaker Feature Engineering Pipelines do	It automates the processes of feature transformation, normalization, and extraction for model training.
What does SageMaker Model Monitor do	It monitors deployed models for performance and data drift.
What is a built-in algorithm in SageMaker	Predefined machine learning algorithms provided by SageMaker for common use cases such as regression, classification, and recommendation.
What is a feature store	A centralized repository for storing, sharing, and managing features used in machine learning models.
What is a SageMaker notebook instance	An environment for building, training, and deploying machine learning models interactively using Jupyter notebooks.
What is a SageMaker training job	A process where data and script are used to train a machine learning model.
What is Amazon SageMaker	A fully managed service for building, training, and deploying machine learning models.
What is AWS Glue	A service used for preparing and transforming data for machine learning.
What is batch training in SageMaker	A method to train large datasets by processing them in batches to manage memory.
What is Elastic Inference in SageMaker	A feature that allows you to attach just the right amount of GPU resources to your inference endpoint, reducing cost.
What is Model Hosting in SageMaker	Deploying trained models as endpoints for inference.
What is reinforcement learning in SageMaker	A method of training models to make a sequence of decisions by interacting with the environment to maximize rewards.
What is SageMaker Autopilot	A service that automates the process of training and tuning machine learning models.
What is SageMaker Experiments	A tool for tracking and comparing the results of machine learning training jobs.
What is SageMaker Pipelines	A workflow orchestration service to automate machine learning workflows, including steps for data preparation, training, and deployment.
What is the benefit of using SageMaker Neo	To optimize machine learning models for deployment on edge devices.
What is the purpose of a validation set in model training	To evaluate the performance of a machine learning model while tuning hyperparameters.
What is the purpose of hyperparameter tuning in ML	To optimize model performance by systematically adjusting model parameters.
What is transfer learning	A method of reusing pre-trained models for a new, similar task.
Why use SageMaker Debugger	To monitor and debug machine learning model training issues.
Why use Spot Instances for training in SageMaker	To reduce the cost of training machine learning models.

AWS Machine Learning Engineer Associate MLA-C01

This deck explains how to train and optimize machine learning models in AWS using SageMaker and other ML-related services.

Share on...