Kaggle, Kickstarter of Machine Learning world, is fixing how we treat heart disease

Written by baris_wonders | Published 2016/01/22
Tech Story Tags: machine-learning | artificial-intelligence | medicine | kaggle | heart-disease

TLDRvia the TL;DR App

Kaggle is running another high profile medical machine learning competition to predict heart disease more effectively from MRI scans. The competition is a joint effort between Booz Allen Hamilton and Kaggle and is also supported by NIH (National Institutes of Health).

If Kickstarter is where you get funding for your next cool project, Kaggle is where you get machine learning experts to solve your big data problem. The process is simple:

  1. You prepare your data and define your problem and set a monetary award for the winner.
  2. The members/teams of Kaggle community submits machine learning solutions to your problem.
  3. You pay the Kaggle member/team with the best model, and you get the license to use it forever.

You can also register to be a solver at Kaggle. In this case, you would be competing with other data scientists to solve big data problems.

Since its start in 2010, Kaggle hosted thousands of such competitions with a variety of prizes. Most recently, they are hosting the 2015 Data Science Bowl challenge where the teams are asked to develop an algorithm that can predict heart disease from MRI images. The first place is awarded $125K and there are still 50 plus days to enter (the dateline is March 2016).

Kaggle is bringing the wiz of data scientists and machine learning specialists to the vast field of medicine, and in this case, transforming how we diagnose heart disease.

It will be exciting to see what other problems will be solved by Kaggle community in 2016. Kudos to its founder Anthony Goldbloom that Kaggle has had such a positive impact so far in the world.

References:

  1. http://www.kaggle.com
  2. http://www.economist.com/blogs/babbage/2011/04/incentive_prizes
  3. http://www.theatlantic.com/technology/archive/2013/04/how-kaggle-is-changing-how-we-work/274908/

Published by HackerNoon on 2016/01/22