HTTPS SSH

Applied AI Internal Demo Project

Feature Reduction with SVD

Autumn 2015

This set of Notebooks and scripts comprise the appliedai_svd internal demo project for Applied AI Ltd.

The project demonstrates using Singular Value Decomposition (SVD) for Feature Reduction in Python.

NOTE: this is a Python 3 project

Setup

Git clone the repo to your workspace.

e.g. in Mac OSX terminal:

    workspace$ git clone https://{username}@bitbucket.org/appliedai/appliedai_svd.git
    workspace$ cd appliedai_svd

Setup a virtual environment for Python libraries

Using Anaconda distro,

  1. create a new virtual environment, installing packages from requirements file

    conda create -n appliedai_svd --file requirements_conda_mac
    source activate appliedai_svd
    
  2. install packages only available on binstar (if relevant)

    (manual commands are shown in file requirements_binstar)
    

Then manually:

  1. install remaining packages from PyPi via pip:

    sh pip_install.sh