Machine Learning Bootcamp Setup Instructions

Install Python

Install Anaconda (Python 2.7) from: https://www.continuum.io/downloads This includes python 2.7.9 and the necessary libraries we will be using: "numpy", "matplotlib", "scipy" and "scikit-learn"

Install Packages with pip

Installing required packages using "pip"

Open your terminal and check whether you have the "pip" function installed by typing pip (and enter). If you do not have pip installed, check the link: https://pip.pypa.io/en/latest/installing/ (If installing via the terminal/command line, ensure you are in the directory where you have downloaded the file "get-pip" or if using chrome right-click on the link to download, save to desktop, and simply double click on the executable).

You may need to use sudo pip install (for OSX, *nix, etc) or run your command shell as Administrator (for Windows) to be able to perform the installation of the following individual packages:

(sudo) pip install Plotly
(sudo) pip install keras
(sudo) pip install --upgrade --no-deps git+git://github.com/Theano/Theano.git

Note that if you're on windows you'll need to install some additional packages to use theano:

conda install mingw libpython

in case that theano doesn't work after installing its dependencies make sure you have the latest version.

If you already have any of the previously-mentioned libraries installed, you can update them to a newer version using the syntax:

pip install <package> --upgrade

where <package> can be any of the aforementioned libraries.

Install required packages with Conda

Once more in your terminal run:

conda install opencv

If the opencv/cv2 library does not load and gives you an error while in the previous step, you may want to try this newer version:

conda install -c https://conda.binstar.org/menpo opencv

If you want to install opencv (including cv2) on OSX with Python 3+ (not 2.7) try this:

conda install --channel https://conda.anaconda.org/menpo opencv3
(import cv2)

Install git

Install git if you don't have it: http://git-scm.com/

Sign up for a GitHub

Clone or download the code from the CCA GitHub repository

You can create a copy of the provided code on your local machine by using the "git clone" command on your console:

git clone https://github.com/cambridgecoding/machinelearningbootcamp.git

Alternatively, click on the "Download ZIP" button under https://github.com/cambridgecoding/machinelearningbootcamp

Download the data

Download the following file (may take some time) and place it in the day2 folder of the ....

VGG16 weights (500MB) from https://drive.google.com/file/d/0Bz7KyqmuGsilT0J5dmRCM0ROVHc/view

Finalise the setup

Navigate to the cloned "machinelearningbootcamp" repository on your local machine, and open and run the "load_libraries.ipynb" file. Wait for the pre-fetching of the CIFAR10 dataset to be completed (it may take a while but a progress bar will show you the remaining time) and check whether the libraries have been successfully loaded. To execute the notebook, in your terminal run:

ipython notebook load_libraries.ipynb

You can run the notebook document step-by-step (one cell a time) by pressing shift + enter.
You can run the whole notebook in a single step by clicking on the menu Cell -> Run All.
To restart the kernel (i.e. the computational engine), click on the menu Kernel -> Restart. This can be useful to start over a computation from scratch (e.g. variables are deleted, open files are closed, etc...).
Click on the menu Help -> User Interface Tour for an overview of the Jupyter Notebook App user interface.

mohfadhil/machinelearningbootcamp