ML Sprouts

Table of Contents

1. Getting Started
2. Charting Basics
3. Comprehensive Overview of Traditional ML
4. Visualizing Data Science Projects with Streamlit
5. Getting More Sample Datasets
6. Miscellaneous Interesting Exercises

Getting Started

ML Sprouts is intended to provide machine learning "sprouts", i.e. a project you can use to select "spouts" to transplant into your ML "garden" to "grow" a new project. Its secondary objective is to provide a comprehensive survey of traditional machine learning methods, new deep learning tools, and a collection of mostly toy examples using small datasets to speed things up.

Most of survey of traditional machine learning methods is extracted from the book "Building Machine Learning and Deep Learning Models on Google Cloud Platform". Note that, in general training deep learning models works best on a GPU instance. There are multiple alternatives for setting up a remote GPU. Some are free, some you will have to pay for, usually by the minute. Prices on services like Paperspace are quite reasonable.

NOTE: The O'Reilly link (below), and links in the code, point to pages in an e-book on their site. The book is also available on Amazon

Contributing: Please create a feature branch before committing code and do a PR when you're ready to merge into the '''develop``` branch. Delete your feature branch after the merge.

Please comment you code and, if you're writing a Python script, add docstrings.

I use Anaconda which will probably load a bunch of packages you won't need but I'm lazy. All the the packages are listed in requirements.txt. If you find you are missing a package that's essential to running your code please add it to requirements.txt. I strongly recommend you create a virtual environment for any project you're working on, even this one. My virtual environment (so I don't forget) is ml

Charting Basics

Comprehensive Overview of Traditional ML

Building Machine Learning and Deep Learning Models on Google Cloud Platform

Visualizing Data Science Projects with Streamlit

Seaborn documentation

Getting More Sample Datasets

Miscellaneous Interesting Exercises

Evolution of a salesman: A complete genetic algorithm tutorial for Python

From the Medium blog post of the same name.