[AIT] MOOC on Algorithmic Information & AI

10 views
Skip to first unread message

Jean-Louis Dessalles

unread,
Apr 25, 2024, 10:51:12 AMApr 25
to ai...@googlegroups.com


Hi everyone,

The MOOC on Algorithmic Information & AI  is back online on EdX.
The current session will run until end 2025.
This is a self-paced course. Access is free, but students have to pay 55€ to get the certificate.

Please encourage your students to enroll.

Thanks,

jean-louis Dessalles


========================================

Dear colleagues,

The  MOOC "AIAI" offers a gentle introduction to Algorithmic Information and its applications to AI.
It's a great suggestion to give to your students before the end of the semester
and also a nice way to motivate them to follow Algorithmic information Theory (AIT) classes
and to question the theoretical foundations of machine learning.

  • Title:
    Understanding Artificial Intelligence through Algorithmic Information Theory
  • Main purpose:
    to bridge the gap between theory and applications
    Also: to make more people aware of AIT and of its importance
  • Links:
    -> see the Presentation of the MOOC
    -> access to the MOOC on the EdX platform

We did our best to find a compromise between rigor and accessibility.
You might consider the MOOC as a tentative popularization of AIT for
students having a bachelor degree in math, computer science, engineering...

The MOOC consists of 5 chapters:

  1. Describing data through compression (code length, Kolmogorov complexity)
  2. Algorithmic Information and Data (complexity and frequency, 'Google' distance, Zipf's law)
  3. Algorithmic Information and mathematics (Algorithmic probability, randomness, Gödel's theorem in three lines)
  4. Algorithmic Information and machine learning (Induction, minimum description length, Analogy, ML as compression)
  5. Algorithmic Information and cognitive AI (subjective information, subjective probability, relevance)
I hope you'll be curious about the MOOC's content,
and that you will welcome its existence.

When applicable, please recommend the MOOC to your students and friends.
And your advice on how to improve the MOOC will be appreciated!


        jean-louis Dessalles


    

    Visit this MOOC: Understanding AI through Algorithmic Information Theory.

            100

    "A very interesting course that definitely changed how I see the world. Thank you so muchhh." (Hanady, 2021, betatester)
    "Very interesting course. It opens a lot of prospects. Thank you very much." (BORG_92, MOOC user)

    "Enlightening. Thank you for organizing this amazing course. I learned a lot! [...]
    After completing the course I’m now contemplating the final chapter on complexity+cognition.
    This chapter was so intuitive and provided such elegant explanations of human behavior."    (Palmaya, MOOC user)


Reply all
Reply to author
Forward
0 new messages