Accelerate data preparation with Amazon SageMaker Data Wrangler
summary
Cloud tool demonstrates a visual data prep workflow that connects to Athena Redshift S3 and SageMaker Feature Store and runs queries and joins without hand coding while providing over 300 transforms for missing values outliers feature generation and string/date cleaning. It previews models by training a random forest with a 70-30 train-test split and shows feature importance and a quick model f1 48.2. Export options produce a notebook or ML Pipeline and can export to Feature Store for production.
AI-written summary · about
chapters (24)
- 0:00Introduction to SageMaker Data Wrangler
- 1:07What is data wrangling?
- 2:13Challenges in data preparation
- 2:56Introducing Data Wrangler features
- 3:39Data selection with Data Wrangler
- 4:23Data transformation capabilities
- 5:05Data visualization and bias detection
- 5:48Quick model performance preview
- 6:54Exporting to production
- 7:59Live demo overview
- 8:43Demo setup and data import
- 10:14Exploring data in Athena
- 13:04Joining datasets from S3
- 16:10Feature engineering with transforms
- 17:11Handling missing values
- 19:23Detecting and trimming outliers
- 20:38Creating binary indicator variables
- 22:28Visualizing data distributions
- 23:39Analyzing delays versus time
- 24:49Subsetting data for modeling
- 26:36Quick model training and evaluation
- 28:09Exporting to SageMaker jobs
- 30:45Customer success and partnerships
- 32:28Getting started with Data Wrangler
related talks
58mPrepare data and model features for ML with ease, speed, and accuracy
44mExploratory data analysis and automated feature engineering
34mPrepare data for machine learning with ease, speed, and accuracy
44mSimplify Data Preparation for Machine Learning
21mAWS On Air ft. Amazon SageMaker Canvas
39mDevelop your ML project with Amazon SageMaker
29mBuilding with Amazon SageMaker Unified Studio
33mBuild, train, and deploy ML models using Amazon SageMaker