Child pages
  • Introduction to Machine Learning (ML) with Python (March 31, 2015)
Skip to end of metadata
Go to start of metadata

When and Where

Date and time: March 31 (5 - 7 PM)
Place: Room 330,  60 Oxford Street, Cambridge, MA 02138

Speaker: K. Indireshkumar (Kumar)

                 Computational Scientist, SEAS Computing

Please note we are using python 2.7 in this tutorial. This is an advanced tutorial introducing some ML tools available in Python and it is  primarily aimed at students in various Harvard classes.  No theory behind various ML techniques will be covered. Background in python, linear algebra, and some basic statistics expected. 

Python Installation (needed only if you want to follow along)

1) Download and install the Anaconda distribution:

     http://continuum.io/downloads

2)  On windows, it should appear in the Start Menu as soon as Anaconda is installed. You can also access the binaries from the command prompt.

     ON Mac, there should be a launcher on the Desktop.

     On Linux (and, perhaps on Mac), you may need to include the path to the Anaconda bin directory in your .bashrc file. It will look something like,

     export PATH="<path_to_anaconda_bin>:$PATH"

     If Anaconda is installed in /home/johndoe/anaconda, then this will be:

       export PATH="/home/johndoe/anaconda/bin:$PATH"

3)  Helpful links on Anaconda, ipython notebooks etc:

        http://docs.continuum.io/anaconda/faq.html

        http://nbviewer.ipython.org/github/catherinedevlin/mpwfw_exercises/blob/master/setup.windows.ipynb

        http://opentechschool.github.io/python-data-intro/core/notebook.html

Tutorial Materials

The Ipython Notebooks for this tutorial are now available (see under downloads below)

The rest of this page assumes you have installed Anaconda and the various python binaries are available in a terminal (Mac and Linux) or command prompt (Windows). Download the Ipython Notebooks below. If you download the notebooks, put them in the same directory.

On Windows, the simplest option is to put these files in the "Ipython Notebooks" directory in the "My Documents" directory under "Documents" (i.e Documents --> My Documents --> IPython Notebooks).

You can open Ipython Notebook on various OSs as follows:

On windows, start the ipython notebook using the launcher under Anaconda in the Start Menu. 

On Mac:

Double click on the launcher (should be available on the desktop) and choose ipython notebook. In the notebook, navigate to the folder which contains the tutorial notebooks.

On LInux, open a terminal (and on Windows, open a command prompt) and change to the directory where your notebooks are. Then type:

ipython notebook

from the directory where you have all the following files (you need to have Anaconda bin directory in the path). Once the ipython server and the browser are up, you will see the files with 'ipynb' extension in the dashboard. Clicking on it will open it.

Downloads for the tutorial:

SVM-activity-identification-03302015.ipynb

PCA-image-compression-03302015.ipynb

HMM-simple-example-03302015.ipynb

Image for PCA:

07122012_EDU_hm2_jfk_main.jpg

Data for SVM:

1.csv

custom.css — for changing the appearance of the notebook (optional)