Welcome to another End to End Machine Learning course!
In this one we’ll build a custom neural network visualization.
What will I learn?
- Concretely, this course is a great way to build up your Matplotlib skills. We will be leaning hard on the basics and making excursions into its advanced functionality.
- Practically, this course teaches you how to build a picture of a neural network. If you research or publish neural networks, this can make your work more visible and easier to explain and debug.
- More generally, you’ll get experience building a custom visualization from scratch. This is a great way to make your analysis stand out or to wrap your head around a complicated algorithm.
How will it work?
The project is broken up into 33 separate coding exercises. In each one I’ll explain to goal, turn you loose to write your own code, and then circle back to talk through how I did it. They range from one line of code to a couple dozen, from straightforward to challenging, but I try to make sure you have all the information you need to do each one.
What do I need to get started?
This course is in python, so some experience writing and running python will be helpful. If you are familiar with functions, for loops, and f-strings you’ll be fine. If all this is new to you, you will still be fine. You will just have to spend a little bit more time Googling and trying things out.
Some things you will not need: previous experience with Matplotlib, data analysis, machine learning, or any understanding of neural networks.
You can use whatever development environment feels most natural to you. I favor writing my python scripts in a text editor and executing them from the command line, but with minor modifications, you can complete this entire course in a Jupyter notebook. Use what is most comfortable.
I’m happy that you’re joining me for this project. Let’s jump to the next lecture where we get more specific about what we want to build .