Score Board

Team ID Team Name Institute Name Country Conservative Liberal Mean Rank Misclassification Error Conservative (%age) Misclassification Error Liberal (%age) Cross-Score Conservative Cross-Score Liberal
Score Rank Score Rank
1327700102 LNCC Laboratório Nacional de Computação Científica Brazil -142 1 -134 3 2 33.57 33.21 -189 -133
1326554855 Soulastral ICFAI India -151 2 -121 2 2 30 31.48 -171 -108
1326396844 DataMiners IIT Kanpur India -153 4 -111 1 2.5 28.83 30.65 -153 -110
1327485937 Lusana Ghent University Belgium -152 3 -168 7 5 37.23 37.22 -152 -182
1328524589 Maverick IIIT Allahabad India -199 7 -161 5 6 35.40 35.40 -199 -161
1327420455 Omega IIT Guwahati India -170 5 -177 8 6.5 35.40 34.30 -174 -186
1328338665 Sublime IIIT Allahabad India -190 6 -190 9 7.5 40.87 39.05 -176 -207
1327080294 FUSION IIT Guwahati India -245 9 -165 6 7.5 45.98 36.86 -198 -282
1328551119 IIITA IIIT Allahabad India -280 11 -152 4 7.5 61.67 36.49 -198 -364
1326533327 Gilnov_jetlui Jaypee University Of Engineering and Technology India -237 8 -194 10 9 41.24 39.41 -218 -225
1326655984 NITK Machinelearners NIT Karnataka India -276 10 -244 11 10.5 69.70 52.18 -214 -429
1327935355 Vigilant Bhoj Reddy Engineering College for Women India -301 12 -326 12 12 76.27 62.40 -236 -495
1326463866 ML National University of Sciences and Technology Pakistan -356 13 -506 14 13.5 69.34 69.34 -356 -506
1326706980 Impregnable Thinkers Vemana Institute Of Technology Bangalore India -384 14 -499 13 13.5 81.75 68.61 -306 -688
1326961346 IITD_SOFT IIT Delhi India -451 15 -687 15 15 82.11 82.11 -451 -687


*The cross score is calculate by interchanging the filters, i.e., applying Liberal filter on Conservative error matrix or vice versa.

The first prize of USD 2000 goes to LNCC, which has the best mean error rank, an interpretable model, which is human-readable. They solve all aspects of the problem.

The second prize is tied between Soulastral, DataMiners and Lusana. All these three rank immediately after LNCC as far as error is concerned. Whereas neither Lusana, nor Soulastral or DataMiners, do anything regarding the model interpretability, but DataMiners and Lusana at least derive insights through selected features. Soulastral does better on error among these three; Lusana has the best and more detailed methodology. All these get USD 660 each. (A three way split for the second prize)

Maverick and Omega did not qualify our benchmark of quality of the methodology as reported in their documents. However as an encouragement, we provide USD 50 each to them.

All the remaining entries get a free machine learning book as a token of appreciation!

Congratulations all!


PS:
  • We will share more details of solutions and prizes soon.
  • We promised total prize money of $5000. We did not get enough entries with good quality that makes these prizes deserving. Whereas we have given out the lion's share of the prize money, we have saved some to spread awareness and education about computer science and social problems among youth over the next one year. We shall share the exact details soon.
  • The results have been arrived upon by the consensus of the experts at MIT and Aspiring Minds after considerable brainstorming. These shall be considered as final and are not subject to any change.