Transform semi-structured nested JSON data for machine learning with no-code solutions on AWS (Level 200)

In many industries, data comes from various sources in structured, semi-structured, and unstructured formats. For semi-structured data, one of the most common lightweight file format is JSON. However, due to the complex nature of JSON data type, it often includes nested key-value structure and is difficult to be used directly in ML tasks. In this session, we discuss how to leverage AWS Glue DataBrew to unnest the data, handle sensitive information, and ensure data quality for ML data preparation. We share how to use Amazon SageMaker no-code solution to automatically train ML models with the processed data to unlock actionable insights quickly. Download slides »
Speakers: 
Melanie Li, Senior Technical Account Manager, AI/ML, AWS
Partha Sarathi Sahoo, Senior Technical Account Manager, Analytics, AWS

Duration: 30mins

Previous Video
Getting started with reinforcement learning and AWS DeepRacer (Level 200)
Getting started with reinforcement learning and AWS DeepRacer (Level 200)

Looking for an interesting and fun way to learn about Reinforcement Learning (RL), then look no further tha...

Next Video
Build accurate models combining diverse data types with AutoGluon on Amazon SageMaker (Level 300)
Build accurate models combining diverse data types with AutoGluon on Amazon SageMaker (Level 300)

Real-world machine learning use-cases often involve data in many forms. In this session, we cover the overv...