Machine Learning at Scale
Joe Minichino · 2023
This chapter explores the various solutions that Amazon Web Services offers in the field of machine learning (ML) and artificial intelligence, with a strong focus on Amazon SageMaker, since it is a product that offers ultimate flexibility. It defines machine learning operations and its role in a data platform. There are several types of ML models: forecasting, regression, classification and recommendation. ML algorithms can be divided into a few families: supervised learning, unsupervised learning, and reinforcement learning. The chapter presents a list that contains a few that have gained particular popularity, but this list is by no means comprehensive: codeguru, comprehend, devops guru, forecast, fraud detector, lex, panorama, personalize, polly, rekognition, sagemaker, textract, transcribe and translate. Inference refers to the predictions made by a machine learning model. There are several ways to perform inference in SageMaker: real time, asynchronous, serverless and batch transform.