There is a strong need to estimate the 90th percentile of the distribution of an environmental variable at 125,480 different locations at 240 distinct points in time. It’s prohibitively expensive and logistically very difficult to make all 125,480 x 240 = 30,115,200 measurements. There has to be a better way which is cheaper and faster but arrives at the same conclusion!
We believe there is, and want you to help us devise an optimal sampling algorithm for this task. With some useful hints and clear constraints, such an algorithm will be valuable and game changing.
To add to the excitement we are sweetening the pot by offering the top 5 winning solutions an opportunity to win a ‘bonus’ prize for the individual whose score has the least variability in results (i.e. least standard deviation) when run multiple times over different tests.
So the question is, are you clever enough to create this algorithm?