Overview
Explore the end-to-end machine learning process using MLflow on Databricks in this 25-minute tutorial. Learn how to leverage health data to predict life expectancy through a comprehensive workflow. Begin with data engineering in Apache Spark, followed by data exploration and model tuning using hyperopt and MLflow. Discover how to utilize the model registry for governing model promotion and deploy models to production as jobs or REST endpoints. Gain insights into the latest innovations from MLflow 1.12, including data cleaning, exploration, modeling, tuning, and production deployment. Follow along as the tutorial covers topics such as Delta Tables, transactional data, exploratory data analysis with Koalas, model staging, and generating predicted life expectancies.
Syllabus
Introduction
Demo Overview
Data Engineering
Data Modeling
Delta Tables
Transactional Data
Exploratory Data
Koalas
Models
Hyperopt
Logging
Model Staging
Model Production
Predicted Life Expectancies
Recap
Taught by
Databricks