Autor Tema: NASA $5,000 Reward  (Leído 3522 veces)

Jack Bauer

  • Administrator
  • Analista
  • *****
  • Mensajes: 237
NASA $5,000 Reward
« en: Abril 19, 2014, 04:16:45 pm »

The best 5 performers of this contest (according to system test results) will receive the following prizes:

    1st place: $5,000
    2nd place: $2,000
    3rd place: $1,500
    4th place: $1,000
    5th place: $500

New topcoder Marathon Match sponsored by NASA with $10,000 in prizes has begun! You have until May 2, 2014 at 13:00 UTC -4 to finalize your submission.

Life as we know it may very well change at any moment for ever without us even realizing it... but don't despair, NASA and Planetary Resources have spun up their engines and they have a plan. When we think of asteroids we immediately think of extinction. It also doesn't help if every once in a while we are reminded of how great that danger really is. Recently the 20m Chelyabinsk asteroid deposited energy equivalent to half a megaton of TNT in the atmosphere over a sparsely populated area, yet still managing to hurt over 1500 people. But aside their destructive force, they also have a brighter side to the story, they may very well be an indispensable mining resource in helping mankind in it's future universe exploration missions!

The Catalina Sky Survey has an algorithm for detecting asteroids and Near Earth Objects (NEOs), but it has an estimated accuracy of 90%, where most of the time it also produces false detections along with the good ones. Increasing the detection efficiency would also increase the number of false detections. Although a trained astronomer could screen out these false detections, this is becoming highly inefficient considering we live in a world with ever increasing data flow from modern instruments. As such, they are looking to employ the minds of ordinary citizens who, although may or may not be trained in astronomy, have the potential of bringing out-of-the-box ideas for developing a post-processing routine that would help reduce false detections, thus significantly improving the detection precision of NEOs.

As part of a joint effort for improving Near Earth Objects (NEOs) detection you are tasked with increasing the accuracy of an algorithm by means of a post-processing step in which your own implementation should properly and automatically screen false detections from the former step thus effectively removing the human interaction factor. For further details, check out the AsteroidRejector problem statement here:

Register to participate in this competition in the Active Contests list here:

For more contest details, click here:

Match forum can be accessed here:

Best of luck to you!

- the topcoder community team
« Última modificación: Abril 19, 2014, 04:18:41 pm por Jack Bauer »