First of all, thank you so much to Jeremy Howard and Rachel Thomas for this fantastic course! After looking around to learn advanced deep learning techniques, I have to say fast.ai is the finest course I’ve ever taken on deep learning.
I am a junior high student currently residing in Pittsburgh, Pennsylvania. My story is that I wanted to look for ways to contribute to the medical field - therefore I started taking Andrew Ng’s Stanford Machine Learning and deeplearning.ai, during the summer after freshman year. While both had vast amounts of theory, it wasn’t enough to be competitive in this continually changing world. I have been since working with a professor in Canada on a medical signal processing project - and I couldn’t have done it without taking this course.
Are there other high school students that would like to share your stories?
James, I’m not a student, but I am a high school administrator. I’m brand new to my school, which is a STEM school in Georgia, and like you, I’m a big fan of fastai. I’m very interested in introducing our students to machine learning and deep learning. However, most do not yet have python or swift programming skills.
I’m curious about a couple of things. Perhaps you can help. I’m wondering what would be the best & quickest way to develop python programming skills among students. I learned initially through Udacity. I know Codecademy is popular. What would you recommend? We have computer technology courses here, but unfortunately, the standards don’t emphasize data science concepts.
Also, and this is probably a question for @jeremy or @rachel, should I focus on exposing students to swift instead? I assume that I should start them on python, then introduce them to fastai, and then introduce them to swift, but I’m not certain.
You should definitely stick with Python for now. There’s a reason the swift material is in lessons 13 & 14, not 1 & 2 - it’s advanced stuff and it’s for researchers interested in what will be the next generation.
I’d love to hear any experiences about teaching high school students!
@travis consider introducing the students to Computer Science Circles from the University of Waterloo in Ontario, Canada. This will take them about a month to complete; once they’ve done that, introduce them to Stanford Machine Learning, Deeplearning.ai and finally Fast.ai. Also, I’d recommend students to learn to use Pandas, Matplotlib, etc. and some more advanced Python techniques (which can be found in the blog Python Tips)
Excellent! Do you teach computer technology, or some other subject? I’d be interested in picking your brain about how you guys implement the STEM concept at your school, and I’d be interested in any suggestions you might have about introducing students to python, data science, and fastai.
Presented an overview of Kaggle and Fastai to some freshman research classes this week. Students were intrigued. Very few had a clear notion of what deep learning is. None had sufficient Python skill to start with Fast.ai right now, but a handful are going to explore the possibility. I have also recruited a senior to team up with me to refine my NCAA basketball tournament predictor, a tabular model.
If the students want to start programming, I think they would benefit greatly from the Waterloo CS Circles. These were great problem-solving resources for me when I started programming.