Machine Learning for Software Developers

This two-day workshop is intended for software developers who have little to no experience in machine learning.

Setting up (do this before the workshop)

The following instructions will guide you through setting up so that you have all the software and datasets needed for the workshop.

Create an account on Kaggle

You'll need this account for some of the projects we'll be doing.

Download the datasets

• Arabic Handwritten Digits Dataset
• MNIST Dataset (download both train.csv and test.csv)

Set up development environment (Windows)

1) Clone this repository.

2) Download and install Anaconda (Python 2.7) from here.

3) Set up the development environment.

To set up the development environment, you'll need to do the following:

• Open Anaconda Prompt

• Run conda create -n ml-workshop. This creates a sandbox where we'll install all the software we'll need.

• Activate the environment: activate ml-workshop

• Install the needed libraries:

  • conda install matplotlib numpy pandas jupyter theano
  • conda install -c conda-forge keras

• Configure Keras (the library we'll use to do machine learning:

  • From your user directory go to the .keras directory.
  • If keras.json isn't present in the folder, create it. If it's there, update it to make it look like this:

  {
    "floatx": "float32",
    "epsilon": 1e-07,
    "backend": "theano",
    "image_dim_ordering": "th"
  }
  

Set up development environment (Mac)

1) Clone this repository.

2) Install Python

Go to your terminal and type python. If you get command not found that means Python isn't installed. You can install it by doing the following:

• Install brew by entering: /usr/bin/ruby -e "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/master/install)"
• Type brew install python

If all goes well you should type python in your terminal and see something like this:

Python 2.7.10 (default, Feb  7 2017, 00:08:15)
[GCC 4.2.1 Compatible Apple LLVM 8.0.0 (clang-800.0.34)] on darwin
Type "help", "copyright", "credits" or "license" for more information.

3) Install Virtualenv

Virtualenv is a python library that helps us build an isolated (virtual) development environment.

• From your commandline, install pip by running the following command: sudo easy_install pip
• From your commandline, install virtualenv by running the following command: sudo pip install virtualenv
• Move to the directory of this repository: cd machine-learning-workshop
• Create the virtual environment: virtualenv ml-workshop
• Activate the environment: source ml-workshop/bin/activate

4) Install required libraries.

After you've activated virtualenv, run the following command:

pip install -r pip-requirements.txt

The command above should install all the software we'll need for the workshop.

5) Configure Configure Keras

Keras is the library we'll use to do machine learning.

• cd ~/.keras
• If keras.json isn't present in the folder, create it. If it's there, update it to make it look like this:

{
  "floatx": "float32",
  "epsilon": 1e-07,
  "backend": "theano",
  "image_dim_ordering": "th"
}