McKinsey Analytics Hackathon - Healthcare Analytics (14th April 2018)

machine_learning
healthcare

#1

Are you ready to crack a real-life business problem on McKinsey Analytics Online Hackathon? April 14th is your opportunity to hack your way into the future—all in a limited span of 24 hours. There is a chance to win an all-expenses paid trip to an international analytics conference of your choice along with interview opportunities.

You can register for the hackathon at this link.

1920x480_VFWdEsJ-thumbnail-1200x1200

The problem is based on a real world problem within the domain of healthcare analytics. The contest starts Saturday, 14th April 2018, 00:00:00 (GMT). The data set and the problem statement will be released at the same time.


#2

#3

Can the leaders share some approaches of attaining 85+ percent accuracies… Thanks


#4

Hi, I am getting an error as forbidden while trying to download the test and train datasets. Can you please help me immediately.


#5

The contest is over.


#6

Hello! This is my first hackathon. I had submitted my solution and marked it as final solution also. But my name is not showing in leader board. I don’t understand what’s wrong. Did the submission not go through?


#7

Wow! Amazing changes between Public and Private leaderboard! People from top positions in public have stepped literateraly into the abyss! Me personally, from Public 52 fell to Private 146. Innocentavali from number 1 to 245!!! Others from top are following the same patterns too! I guess having unlimited access to submissions lead us to overfit our models to the public leaderboard dataset…A big lesson for me (and many others I guess)! Thanks McKinsey and AV!!!


#8

I was at 220 in the Public Leaderboard! After the close of deadline… I am at the 27th position. The focus was on building a model with minimal variance in Local and LB; my local AUC was at 82.2 and Public LB was at 83.


#9

Could you share more details about your approach or code?