Dear all,
Hoping we can still call Wednesday morning the "beginning of the week" :), here is finally some detailed info about tomorrow's competition:
- First, you will need to be connected via vpn (one member per team, at least) or to be on EPFL site in order to upload your solution. Here are the setup instructions, in case you never used the vpn:
- Next, as already said, you will be given two datasets (access to be provided tomorrow), the first one with the order of tens of thousands of cities, the second one with the order of thousands of cities, to optimize each time with a different value of lambda.
- The format of each dataset is a csv file with the following fields in each line:
city id (=number between 1 and n), normalized population, position x, position y
[with x,y being numbers between 0 and 1; the normalized populations might be larger than 1, but not much more than that]
and you should then upload a csv file with the following fields in each line:
- The winning team is the one reaching the highest score on the largest dataset, with the score on the smaller dataset as a second criterion in case of equality. The winner takes it all: automatically full grade for the project, plus 5% bonus on the final grade.
All the best, and see you tomorrow on Zoom (usual course link on the course webpage) !
Olivier
PS: We will create breakout rooms to allow you to discuss in private with your team during the competition.
