#100DaysOfMLCode: A Machine Learning Journey (Summary)

100 days ago, AI Youtuber Siraj Raval proposed the “100 Days of ML Code” challenge. In this challenge you pledge to code and/or study machine learning for at least an hour everyday for 100 days. You also keep a log of your progress through social media.

As a ML freak with time on his hands, it felt natural to accept the challenge. Twitter was the platform of choice. What I discovered is that Twitter gives you a lot of opportunity to interact with people that you don’t know personally. For example, Nikhil Seth made a Twitter bot that retweeted every post with #100DaysOfMLCode. This really enhanced the interaction and experience for everyone doing the challenge. Furthermore, a lot of top AI researchers like Jeremy Howard, Francois Chollet and Ben Goertzel are very active on Twitter!

Anyway, here is my summary for the #100DaysOfMLCode challenge:

3 books (and some partly read books):
“Deep Learning” by Ian Goodfellow, Yoshua Bengio and Aaron Courville
“Superintelligence” by Nick Bostrom
“Think Python” by Allen Downey

6 free online courses:
Machine Learning by Georgia Tech (Udacity)
Machine Learning for Trading (Udacity)
DeepMind’s David Silver Reinforcement Learning Course
Fast.ai’s “Deep Learning for Coders” Part 2 (Heavy stuff!)
School of AI’s Move37 Course (Currently Studying)
Fast.ai’s “Machine Learning for Coders”  (Currently Studying)

4 (serious attempts at) Kaggle competitions and creation of new Kaggle team “Wireless Infidelity”:
PUBG Finish Placement Prediction (Still Running) (Top 2% with Wireless Infidelity)
Digit Recognizer (Top 10%)
Titanic: Machine Learning from Disaster (Top 45%)
Human Protein Atlas Image Classification (Still Running) (Top 60%)
Kaggle Profile: kaggle.com/carlolepelaars

4 Github Repositories created:
bipedalrobot
MiniMNIST
PyMNIST
cost_functions
Github Profile: github.com/CarloLepelaars

1 School of AI Meeting in Enschede, The Netherlands
Video

Lots of Coffee
Lots of Python
Lots of Videos
Lots of Papers
Lots of Math

Feel free to connect on Twitter and talk about AI:
@carlolepelaars

If you like AI courses, videos, books, etc. check out https://aitube.io