Responsible AI Course Free From Amazon
Written by Sue Gee   
Thursday, 05 January 2023

Bias and fairness have become important issues as AI and machine learning continue to take over decisions in all walks of life. Amazon's Machine Learning University has added a new course addressing these issues with practical application that is publicly accessible online for free.

As an organization, Amazon is an important pioneer of AI. Back in 2015 when Amazon Machine Learning was launched, product recommendation based on customer purchase history was used as the main illustration. Since then Amazon's use of ML has diversified but remains at the centre of its business - from supply chain optimization through to the way the Alexa voice assistant interacts with its users. 

Amazon MLU

Amazon founded the Machine Learning University (MLU), in 2016 as an in-house facility in which classes are taught by Amazon ML experts. Its curriculum is designed to:

sharpen the skills of current ML practitioners, while also giving neophytes the tools they need to deploy machine learning for their own projects.

In 2018 Amazon opened MLU up to developers via AWS, see Free Machine Learning Training From Amazon and two years later made it even more accessible by delivering its content via a You Tube channel, as reported in Amazon Makes Machine Learning Courses Available to All.

At that time there were three courses in Natural Language Processing, Computer Vision and Tabular Data, each of them providing lectures on YouTube backed up by slides and Jupyter notebooks with datasets and hands-on exercises on GitHub. The next course to be added, in 2020, was Decision Trees and Ensemble Models.

Now there's a new entry-level course, that as its full title, Responsible AI — Bias Mitigation & Fairness Criteria, makes clear has the goal of explaining where bias in AI systems comes from, how to measure it, and ultimately how to mitigate bias as much as possible.

In-house at Amazon MLU it is presented as a 3-day course and taken online its YouTube playlist comprises 30 videos. 

In-house at Amazon MLU it is presented as a 3-day course and taken online its YouTube playlist comprises 30 videos. 

As explained in the introductory video by Mia Mayer who developed the course, the learning outcomes have two components - theoretical and practical. As regards the former the course provides a fundamental understanding of Machine Learning, covering concepts and terminology. It then moves on to imparting practical ML skills and the techniques used to train, tune, test and evaluate simple ML models. Having set out the basics the course focuses on identifying bias and ways of mitigating it. In addition to the videos there are white papers to read, and Jupyter Notebooks on GitHub with data and code samples. You'll need an AWS account  in order to complete a final project, students implement their own bias mitigation technique of choice to reduce disparity in model outcomes for different subpopulations.  

 

Amazonmluni

 

 

 

More Information

Machine Learning University

Responsible AI Playlist on You Tube

Related Articles

Amazon Makes Machine Learning Courses Available to All

Free Machine Learning Training From Amazon

Amazon's Giant Push Into Machine Learning

Amazon AI Services

Amazon Rekognition Can Now Estimate Your Age

To be informed about new articles on I Programmer, sign up for our weekly newsletter, subscribe to the RSS feed and follow us on Twitter, Facebook or Linkedin.

 

Banner


GitHub Has 100 Million Active Users
27/01/2023

In his keynote at GitHub Universe 2019, then CEO Nat Friedman announced GitHub had more than 40 million active users and predicted that by 2025 it would have 100 million. That target has already been  [ ... ]



GitHub Adds Required Workflows For Actions
16/01/2023

GitHub has introduced required workflows and support for configuration variables for Actions. The update is designed to standardize and enforce CI/CD best practices, and give developers working for la [ ... ]


More News

picobook

 



 

Comments




or email your comment to: comments@i-programmer.info

Last Updated ( Thursday, 05 January 2023 )