|Version 12 (modified by jarrod.millman, 6 years ago)|
Scipy Summer of Code
According to the Google SoC timeline, coding (and payments) start on May 28th. That should be plenty of time to make sure you have everything you need to succeed, and that all infrastructure issues, both with Scipy and with Google, get resolved.
Before May 28th, you should make sure that you have completed all of the "First Steps" listed below:
- Get a subversion account for Scipy and Scikits.
- Join both the Scipy user and developer mailing lists. You should view these lists as a resource and a place to float ideas coming out of your project work. Please make active use of these lists: we will be reading them too, and it gives you access to the resources of a larger community.
- Refine scope, milestones, and deliverables.
- Create a blog for your project. As per the PSF's request, you should post notes about progress, problems, discoveries, etc. All the standard blogs provide syndication feeds in at least one XML format. These XML feeds will be accessed periodically by a script and compiled into a "group blog" (for example, see http://www.planetpython.org/). If you don't have a preference, you should probably just use blogger.com since it's a Google service.
- Install the development version of Scipy from subversion. Make sure to run all the tests.
- Familiarize yourself with the relevant sections of the Scipy code.
- Read through the guidelines below including reading the links.
- Style: Please install pylint and run it often as a check on your code quality.
- Testing: Please note that good documentation strings and unit tests are a crucial part of Scipy code base. These are absolutely necessary for your code to be usable by others.
- License: All Scipy code should be made available under a revised BSD-style license.
MachineLearning for SciPy and SciKits
Student: David Cournapeau Mentor: Jarrod Millman Backup mentor: Brian Hawthorne
Python-based equivalent to commercial modelling systems for optimization problems (AMPL, GAMS, TOMLAB etc)
Student: Dmitrey Kroshko Mentor: Alan Isaac Backup mentor: Anne Archibald
Neural Nets in SciPy
Student: Frederic Mailhot Mentor: Robert Kern
The goal of this project is to extend SciPy's functionality by adding modules for the design, training and use of a variety of neural network architectures, including standard feedforward and recurrent networks, among others. As a guide I intend to work from the modules in Matlab's Netlab toolkit, as well as from my own experience implementing recurrent networks.
Support Vector Machines for SciPy
Student: Albert Strasheim Mentor: Dave Kammeyer
Integrating libsvm (and possibly other SVM libraries) with SciPy using ctypes and providing various tools for manipulating datasets for use with SVMs.