Hello! This is the DeepCourse and I’m Arthur, a PhD student that loves sharing what I’m learning. This course is the new version of a lectures serie I’m giving every Fall at the French engineering school EPITA.
What is this Course?
This course comprises several methodologies that I think are useful for students:
- First head to the lesson (in yellow) and learn the new topics
- Then practice by coding algorithms in the multiple exercise session (in orange)
- Do a larger quiz to see again if you understood well every aspect of the lesson by doing active recall and space-repetition using Anki!
What is Anki?
Whether you’re still in school or not, it’s very probable that you passed several classes by cramming knowledge. It may have provided you a good grade, but you also probably forgot a lot of it.
Could we learn to have a good grade AND not forgetting long-term AND not passing hours per day to review the topics acquired years ago?
- Yes, we can.
The magic resides in space-repetition. Gwern.net has made a great overview of this concept, but to put it shortly it is the act of reviewing a fact when you’re about to forget it. The following image (source) shows the forgetting curve once we memorized the fact for the first time:
Projected forgetting curve and space-repetition
Reviewing several times, at longer interval each time –if you do good, helps to avoid forgetting.
Because the intervals are getting longer gradually, you won’t feel overwhelmed under many cards. For example, I learned a bit more than 8000 cards in various subjects (ranging from math, CS, history, Chinese, geography, arts, etc.) but I rarely have more than a hundred cards to review per day, less than 20min.
Anki is one of the most popular space-repetition software (SRS), mainly because of its vibrant community that shares decks and addons. It is available on Windows, Mac, Linux, Web browsers, android, and iOS. Note that all but the latter are free (although it’s worth it).
I cannot stress enough how space-repetition is amazing. Please try it.
Acknowledgements
I had many inspiration for this course, here are some:
- Polytechnique’s master of Data Science
- This is the course Charles Ollion taught me at EPITA. He then task me to taught this course myself at EPITA. Thus this current course is in some part very similar to it.
- Sorbonne’s RDFIA: a course I’m teaching at Sorbonne. Most of the content was done by Thomas Robert and Rémi Cadene.
- Stanford’s famous CS231n
- Quantum Country which motivated me to make my own course with space-repetition