Duration
14 hours (usually 2 days including breaks)
Requirements
- Experience with machine learning algorithms.
- Python or R programming experience.
Audience
- Data analysts
- Data scientists
- Data engineers
- Developers
Overview
AutoML is user-friendly machine learning software that automates much of the work needed to select an ideal machine learning algorithm, its parameter settings, and pre-processing methods.
This instructor-led, live training (online or onsite) is aimed at technical persons with a background in machine learning who wish to optimize the machine learning models used for detecting complex patterns in big data.
By the end of this training, participants will be able to:
- Install and evaluate various open source AutoML tools (H2O AutoML, auto-sklearn, TPOT, TensorFlow, PyTorch, Auto-Keras, TPOT, Auto-WEKA, etc.)
- Train high quality machine learning models.
- Efficiently solve different types of supervised machine learning problems.
- Write just the necessary code to initiate the automated machine learning process.
Format of the Course
- Interactive lecture and discussion.
- Lots of exercises and practice.
- Hands-on implementation in a live-lab environment.
Course Customization Options
- To request a customized training for this course, please contact us to arrange.
- To learn more about AutoML, please visit: https://www.automl.org/
Course Outline
Introduction
Setting up a Working Environment
Overview of AutoML Features
How AutoML Explores Algorithms
- Gradient Boosting Machines (GBMs), Random Forests, GLMs, etc.
Solving Problems by Use-Case
Solving Problems by Training Data Type
Data Privacy Considerations
Cost Considerations
Preparing Data
Working with Numeric and Categorical Data
- IID tabular data (H2O AutoML, auto-sklearn, TPOT)
Working with Time Dependent Data (Time-Series Data)
Classifying Raw Text
Classifying Raw Image Data
- Deep Learning and Neural Architecture Search (TensorFlow, PyTorch, Auto-Keras, etc.)
Deploying an AutoML Method
A Look at the Algorithms Inside AutoML
Ensembling Different Models Together
Troubleshooting
Summary and Conclusion