2009 IEEE ICDM Data Mining Contest	ICDM 2007 Contest ICDM 2008 Contest
	2009 IEEE lCDM Conference on Data Mining Pittsburgh Brain Competition Brain (PBC) Connectivity Challenge Organized by: Professor Walter Schneider at the University of Pittsburgh Details: www.braincompetition.org

Brief summary:

The competition involves applying data mining to map the "cables of the human brain" mapping ~300,000 fiber streamlines into 20-50 cables or tracts. It involves two challenges:

1. Supervised learning to match the categorization of an expert

2. Unsupervised learning of within and between subject segmentation

   

   Figure 1. Major fiber tracts that are the target for the competition

A short YouTube presentation details the competition including illustrations of the fiber tracts. Details of the competition are available at the web site www.braincompetition.org. All entrants will be able to compare their scores relative to all the other completed entrants. Awards will be announced at the  ICDM 2009 Conference, December 2009.

Awards will include:

• Supervised Learning Match Expert Award $2000 (US);
• Unsupervised Learning Within and Between Subject Award $2000 (US);
• and possibly a Contest Crown Award (if a group scores top in both award categories), including $1000 (US) and an option to have the team leaders have their brain scanned to create a fiber map of their brain (which is expensive!).

• Competition website http://www.braincompetition.org
• Youtube Overview http://www.youtube.com/watch?v=IIDv4FLRg_4
• IEEE Int. Conf. on Data Mining http://www.cs.umbc.edu/ICDM09/
• Competition Full description http://schweb1.lrdc.pitt.edu/pbc/2009b/media/IEEE_IDCM_PBC_BrainCompetitionRules.pdf
• Quesitons: contact pbc (at) pitt.edu

• Training data available: September 14, 2009
• Full data released: Oct 1, 2009.
• Results due November 2, 2009, write-ups due November 6, 2009. 

  This contest is organized and sponsored by the University of Pittsburgh, Professor Walter Schneider, the competition director. 
  Chair for the IEEE ICDM 2009 Data Mining Contest: Qiang Yang.