Three Concepts: Probability, The Projects

Round 1: Results

The Contestants

Here are the participating teams with their scientists:

Team Scientist
Team 3CP-A OurScientist
Team 3CP-B Alpha
Team 3CP-C OmaScientist6
Team 3CP-D stypido.py
Team 3CP-E Sci3
Team 3CP-F NeuralScientist
Team 3CP-G Akvavitix

The Trial

Each scientist was tested against a set of five distributions. Each distribution was tried eleven times independently with different random seeds. Identical seeds were used with all the competitors.

The Results

Here are the results. Listed are the medians of the runs and corresponding standard deviations in the parenthesis. Columns are the distributions having arbitrary names. Now, drums please!

  Data1 Data2 Data3 Data4 Data5
OurScientist 109158 (297115.9) 275 (111.3) 1011 (435.1) 238 (82.4) 285 (163.6)
Alpha 852 (3876.4) 316 (138.2) 670 (450.2) 192 (52.9) 284 (142.1)
OmaScientist6 1000000 (466918.3) 347 (407.3) 1460 (504182.5) 188 (62.3) 1000000 (521929.9)
stypido.py 1000000 (0.0) 69189 (5416.8) 1000000 (288416.1) 6435 (1017.3) 1000000 (0.0)
Sci3 2178 (1146.4) 381 (673.8) 1122 (542.9) 373 (91.9) 404 (143.2)
NeuralScientist 1024 (504291.7) 275 (301427.3) 839 (425.2) 291 (76.0) 261 (368.8)
Akvavitix 1000000 (0.0) 1000000 (301412.7) 1000000 (0.0) 269 (125.5) 1000000 (0.0)

Round 1: Hall of Fame

Here's the ranking after round 1. We ordered the scientists per every distribution by the medians shown above and took the average position over the distributions as the final rank.

Avr. Ranking Scientist
1.8 Alpha
2.2 NeuralScientist
2.8 OurScientist
4.0 OmaScientist6
4.4 Sci3
5.4 Akvavitix
5.8 stypido.py

 

 Three Concepts: Probability
2006