Machine Learning Fundamentals - Understanding and Utilizing the Python Library Stack

This repository hold the notebooks, docker images, exercises, etc. for the Machine Learning Fundamentals.

Prerequisites

1. git - please install and have a basic working knowledge of git.
2. docker - please install and have a basic working knowledge of docker.
3. python - you may want to have Python3.7 installed, although this isn't required. You will want at least an introdctory level of understanding programming in python. (There are numerous sources that will get you up to speed programming python. Try a google search and select the one you fancy.)
4. jupyter - should you elect to install python 3, you may also want to install jupyter, although this is not required. We also recommend watching or reading through a tutorial to glean a basic understanding of Jupyter prior to participating in these sessions (again, google search...).

Python The Hardway

• Learn Python The Hard Way

Other Resources for Understanding Machine Learning

Math Refresher

Linear Algebra

• An Intuitive Guide to Linear Algebra
• The Essensce of Linear Algebra (YouTube)
• MIT OpenCourseWare

Calculus

• The Essensce of Calculus (YouTube)
• MIT OpenCourseWare

Resources for Learning

• Neural Network Playground
• ML Crash Course
• Coursera Machine Learning
• Coursera Deep Learning
• Udacity Machine Learning
• Deep Learning by University of San Francisco
• Python Programming
• Computational Linear Algebra
• Spark

Interesting Podcasts/VideoCasts

• Linear Digressions
• Data Skeptic

Adding python packages to the requirements.txt

• Add to the file docker/root/tmp/requirements.txt
• Please don't add the version so that it makes it easier for pip to pick the latest stable version.

How to use this docker?

• Open windows command prompt or Linux/Mac termrminal.
• Run the following commands in terminal mode
• Just clone the package with the following command

  git clone https://github.com/spendyala/deeplearning-docker.git
  

• It should open a browser with Jupyter Notebook.

  run-notebook
  

  • For the first time it takes longer to download the docker and install and configure the container.
• To build a docker from scratch and start using the following command.

  run-notebook --build
  

  • For the first time it takes longer to download the docker and install and configure the container.

Some of the docker commands

Info	command
Running docker containers.	docker ps
Docker images	docker images
Purging none images to clear space	docker system prune -a
Remove images	docker rmi <image>
SSH into running container	docker exec -it <container_id_from_ps> /bin/sh
Docker Login	docker login
Tag image	docker tag deeplearning vpendyala/deeplearning:latest
Push image	docker push vpendyala/deeplearning:latest

Credits to Daniel Rapp.

Without his help and knowledge, this docker image might have taken more time to consolidate and compose this Dockerfile.