Automated machine learning (AutoML) enables both data scientists and domain experts (with limited machine learning training) to be productive and efficient. In recent years, AutoML has fostered a fundamental shift in how organizations approach machine learning, making it more accessible to both experts and nonexperts.
Most real-world data science projects are time-consuming, resource intensive, and challenging. Besides data preparation, data cleaning, and feature engineering, data scientists often spend a significant amount of time on model selection and tuning of hyperparameters. Automated machine learning changes that, making it easier to build and use machine learning models in the real world.
Francesca Lazzeri and Wee Hyong Tok (Microsoft) lead a gentle introduction to how AutoML works and the state-of-art AutoML capabilities that are available. You’ll learn how to use AutoML to automate selection of machine learning models and automate tuning of hyperparameters.
- An introduction to AutoML
- How AutoML works
- The libraries and cloud services that support AutoML
- An energy demand forecasting use case
- How to get started with automated machine learning
This session was recorded at the 2019 O'Reilly Artificial Intelligence Conference in New York.
- Title: Using AutoML to automate selection of machine learning models and hyperparameters
- Release date: October 2019
- Publisher(s): O'Reilly Media, Inc.
- ISBN: 0636920339588
You might also like
Strata Data Conference 2019 - San Francisco, California
Thousands of the data scientists, analysts, engineers, developers, and executives converged at the Strata Data Conference …
O'Reilly Strata Data Conference 2019 - New York, New York
The 2019 Strata Data Conference NYC, the biggest Big Data conference in the world, was a …
Hands-On Machine Learning with Scikit-Learn, Keras, and TensorFlow, 2nd Edition
Through a series of recent breakthroughs, deep learning has boosted the entire field of machine learning. …
Designing Data-Intensive Applications
Data is at the center of many challenges in system design today. Difficult issues need to …