This repository holds our announcement about future events
Explainable AI study group aims to create a low-pressure, peer-mentoring learning space for students to learn and practice explainable AI (XAI) techniques. For each session, one instructor will walk the audience through a working example of an explainable AI method. Unlike conventional reading groups, this group emphasizes hands-on learning. And learners do not need to prepare before the sessions. You are welcome to “watch” this repository for any updates.
To attend our sessions, you only need to have two types of IDEs (Integrated programming environment) installed on your computer and know how to use them. One for R and one for Python, since we allow instructors to choose which language they want to code their examples in. No prior experience in ML is needed, but having it is a plus.
If you would like to recieve updates on our sessions, please make sure you are a viewer of the “XAI study group documentation” in Box. I use Box’s “mail all” function to update your about upcoming events, and I keep information that I don’t feel comfortable putting on GitHub on the Box documentation (closed membership), such as Zoom session recordings.
To view all the repositories contributed by our instructors, please visit our GitHub Home
Since this is a peer-mentoring learning space, we do not hold instructors to the same high standard for a real classroom. You can choose a method, code a working example of using this method to explain your machine learning or deep learning model, and show how you do it to the audience. You also need to demonstrate how this technique help you to understand your model better.
We welcome any upper-level undergraduate students, master’s students, and PhD students to try on the role of instructors. Develop your code on your own and let me know when the repository is ready to be forked.
If you are not confident about your teaching skills, please check out resources from U of I’s Center for Innovation in Teaching and Learning. If you want to get more experience teaching programming and data science skills, I also recommend joinning the Carpentries as a volunteer.
Open an issue in this repository, so we will see it.
As a learner, you will learn techniques that can be used to explain your own machine learning or deep learning models, which will make you more competitive on the job market when the majority of people only know how to build their models and assess models with standard metrics such as F1 scores. As an instructor, you learn teaching skills, curriculum development skills, and the experience of working with a diverse group of learners.
For each session, one instructor will walk the audience through a working example of an explainable AI method.
Recordings are kept for 14 days, and links are only available to group members.