Augmenting Machine Learning with Databricks Labs AutoML Toolkit

Explore how to streamline and optimize machine learning processes using the Databricks Labs AutoML Toolkit in this 30-minute conference talk. Learn about the challenges data scientists face when creating ML models, including data preparation, feature engineering, model selection, and optimization. Discover how AutoML can significantly simplify these tasks through a demonstration using financial loan risk data. Gain insights into AutoML's tiered API approach, exploratory analysis for feature identification, ML pipeline stages, and model experimentation. Compare hand-made models with AutoML-generated ones, and understand how to interpret metrics and configurations. Delve into common overrides, business value considerations, and the AutoML family of tools. Conclude with a look at the AutoML roadmap and access downloadable code snippets and notebooks to apply these concepts in your own projects.

Intro
About Speaker
AutoML's Tiered API Approach
Let's start at the end
Exploratory Analysis to Identify Features
ML Pipeline Stages
Hand-made Model
Model, Metrics, Configs Saved
How did Auto ML Toolkit do this?
Common Overrides Override
Clearing up the Confusion A
Business Value
Model Experimentation
AutoML FamilyRunner
Let's end at the end
AutoML Roadmap