R(ed)Panda - Machine Learning notes

“For the things we have to learn before we can do them, we learn by doing them.” ― Aristotle, The Nicomachean Ethics

This is a red panda.

In this section I decided to collect some notes on machine learning. There are many good reasons to do so, and for me in particular these are the main ones:

• learn by doing: as great it is to have access to many open-source frameworks to do just about anything I like to spend time writing down simple algorithms from scratch. Nonetheless I am not trying to reinvent the wheel here so I do use a minimal set of tools for numeric computation and linear algebra (mainly counting on the help of NumPy).
• write down the math: it is very easy sometimes to overlook the math behind a fit() call.
• collect practical insights from real data: well, this is all fun but it’s no use if the algorithms are not helping use to actually solve problems.
  

For these reasons I created a R(ed)Panda, a repo to collect some basic implementation of some of the most known machine learning algorithms to play around with and report here whatever seems interesting.

Table of Contents

1. Supervised Machine Learning
   • Theoretical minimum (TODO)
   • Linear regression (TODO)