What is the Kaggle Data Science Bowl 2017 competition about?

Click For Summary

Discussion Overview

The discussion revolves around the Kaggle Data Science Bowl 2017 competition, which focused on creating a model to detect lung cancer. Participants share their thoughts on the competition's structure, its implications for the medical community, and the nature of competitive data science challenges.

Discussion Character

  • Exploratory
  • Debate/contested
  • Technical explanation

Main Points Raised

  • Some participants express enthusiasm for Kaggle competitions, highlighting the opportunity for diverse participants to contribute, regardless of their background.
  • One participant notes the potential benefits for the medical community from advancements in cancer detection models.
  • Another participant shares their experience in a different competition focused on recognizing sea lions, emphasizing the learning experience and competitive nature.
  • There is a mention of a current Kaggle competition on predicting house prices in the Russian housing market, with interest in the methods used by contestants.
  • Participants discuss the incentive of cash prizes as a motivating factor for engagement in these competitions.

Areas of Agreement / Disagreement

Participants generally agree on the value of Kaggle competitions and the potential benefits they bring, particularly in the medical field. However, there are multiple competing views regarding the specific competitions being discussed, such as the lung cancer detection and the sea lion recognition challenges.

Contextual Notes

Some discussions reference the competitive nature of Kaggle challenges and the varying levels of experience among participants, which may influence outcomes and learning opportunities.

Who May Find This Useful

Individuals interested in data science competitions, machine learning applications in healthcare, and those looking to learn from practical challenges in the field may find this discussion relevant.

Jameson
Insights Author
Gold Member
MHB
Messages
4,533
Reaction score
13
There is a neat site called Kaggle that is home to lots of data science info and the place of very featured competitions with large cash prizes. The goal of this year's competition was to create a model to detect lung cancer. It just wrapped up last week and the results are being verified right now as far as I know. $500,000 goes to the 1st place model, so very big money here.

Since I have become a professional model builder/checker, I find this kind of thing really cool and a great idea. Present a challenge to the whole internet with a large cash prize and see what comes back. You could be a grad student, a medical doctor, or completely untrained in the traditional ways but if your model is the best you win.

In a way this is like the Millennium Challenge Problems, each with a $1 million prize but this is more applied and coding centered than those (well perhaps except the P=NP proof :D).

Anyway, what do you guys think about this competition and these kinds of competitions in general?
 
Computer science news on Phys.org
I like it. Particularly, with the focus on cancer, who knows what benefits for the medical community may arise.
 
I am currently participating in a competition to recognize sea lions. The competition is fierce to be honest with many previous winners but it is always a good experience that you learn from. Also the Cash prize is a good incentive.
 
Yep. Nothing like a bit of money to force one to pull his finger out.
 
ZaidAlyafey said:
I am currently participating in a competition to recognize sea lions. The competition is fierce to be honest with many previous winners but it is always a good experience that you learn from. Also the Cash prize is a good incentive.

What contest is this for?

Currently there is another Kaggle competition on predicting house prices in the Russian housing market. I have done a similar project for my actual job so am very curious on the methods that contestants use.
 
Jameson said:
What contest is this for?

Currently there is another Kaggle competition on predicting house prices in the Russian housing market. I have done a similar project for my actual job so am very curious on the methods that contestants use.

It is a computer vision competition . The question is, given an image how many sea lions are there ? Actually it is more complicated since it asks for the type as well.