/math-app-ml

Essential mathematics for applied machine learning and data science

Primary LanguageJupyter Notebook

Essential Mathematics for Applied Machine Learning and Data Science

What is this?

A collection of interactive tutorials about essential mathematics for applied machine learning and data science.

How are you doing this?

As a learning resource with the following characteristics:

  • Open
  • Free
  • Interactive (Jupyter Notebooks and blogpost formats)
  • Visual
  • Python-based
  • Math with code, i.e., exemplifying mathematical concepts with Python.

Is this for me?

Maybe. I have no formal mathematical training beyond high school, so I'm writing this from that perspective. This entails the following:

  1. I use a narrative style, like a conversation with myself or a dear friend who I'm trying to help to understand something.
  2. I try to explain every bit of notation and to "translate" math expression into natural language.
  3. I use visual explanations and analogies as much as possible.
  4. Almost zero proofs.

What are the prerequisites?

This is a tricky question. I'd like to say "NONE! Only your enthusiasm!" but I would be lying. That being said, my best guess is: solid high school math. Now, from where I come from, math education is pretty bad, therefore is unclear what solid means for people from different parts of the world. From my experience writing this I'd say:

  1. Algebra
  2. Pre-calculus
  3. A bit of trigonometry
  4. Basic notions of probability and sets

What contents do you cover?

  1. Linear Algebra progress
  2. Calculus (Differential, Integral, and Vector Calculus) progress
  3. Probability and Statistics progress
  4. Information Theory (maybe)
  5. Optimization (maybe)

Points (4) and (5) are maybes, not because they aren't important, but because my time availability.

How do I use this?

You have three options:

My blog

Read the tutorials in my blog.

Remotely

With mybinder, by clicking the icon -> Binder

This will build a docker image of the repo that runs on the cloud. Beware that it may take 2-3 minutes to be ready. Note the mybinder link may be behind a few updates.

Then navigate to the notebooks directory and access to the tutorial.

Locally

To obtain the files locally, run this in the command line:

git clone https://github.com/pabloinsente/math-app-ml.git

To set up your system, you need python 3.6.x. It is recommended to use a virtual environment before installing the dependencies. To do this, navigate into the cloned repository in the console by:

cd math-app-ml

Note that you may need to change the path to cd into the directory.

Then run this inside that directory to create the virtual environment:

python3 -m venv venv

And activate your environment by running:

source venv/bin/activate

Make sure to have the latest pip version:

pip install --upgrade pip

Install dependencies by running:

pip install -r requirements.txt

To run the notebooks, navigate to the notebooks directory and launch Jupyter Lab as:

jupyter lab

Who are you and why are you doing this?

I'm PhD student in Psychology at UW-Madison. I'm from Santiago, Chile. I'm doing this for a couple of reasons:

  1. Because teaching (or creating educational content) is the best way to learn for me. I've TA multiple courses over the years, and as a result became convinced the person learns the most after each class is the instructor, not the students.
  2. Because I like to do things that other people can use, even if is just a small number of individuals. Particularly if those individuals are people like me, that sometimes do not have the money to pay for books or courses.
  3. Because is a way to show potential employees that I know "things", even math, which is something most people don't believe given my lack of formal training.