ECE 8527: Final Exam - Results of the Final Exam Data Challenge (Set 14)


The data set for this challenge can be found here. Note that there are five classes (background plus sets 1-4). The scoring metric consisted of:

      P = average[TPR for C1-C4] - average[FPR for C1-C4]

A summary of the results in spreadsheet form can be found here. Assuming you solve the segmentation problem, random guessing results in an accuracy of 25% (forced choice on the four non-background classes).

Participant Algorithm Train Dev Eval
SPRING 2022
Kane, Zach (2022 Spring) K-MEANS (KMN) 30.98% 30.34% 38.94%
Kane, Zach (2022 Spring) Autoencoder (AUT) 29.86% 29.47% 32.45%
Samarco, Michael (2022 Spring) ConvNet (CNN) 60.66% 60.07% 60.61%
Samarco, Michael (2022 Spring) Random Forest (RNF) 0.00% 0.00% 0.00%
Bici, Daniel (2022 Spring) TBD (TBD) 0.00% 0.00% 0.00%
Bici, Daniel (2022 Spring) TBD (TBD) 0.00% 0.00% 0.00%
Cassell, Joshua (2022 Spring) TBD (TBD) 0.00% 0.00% 0.00%
Cassell, Joshua (2022 Spring) TBD (TBD) 0.00% 0.00% 0.00%
Khantan, Mehdi (2022 Spring) TBD (TBD) 0.00% 0.00% 0.00%
Khantan, Mehdi (2022 Spring) TBD (TBD) 0.00% 0.00% 0.00%
Rahman, Nazia (2022 Spring) TBD (TBD) 0.00% 0.00% 0.00%
Rahman, Nazia (2022 Spring) TBD (TBD) 0.00% 0.00% 0.00%
Sand, Richard (2022 Spring) TBD (TBD) 0.00% 0.00% 0.00%
Sand, Richard (2022 Spring) TBD (TBD) 0.00% 0.00% 0.00%
Vadimsky, Dakota (2022 Spring) TBD (TBD) 0.00% 0.00% 0.00%
Vadimsky, Dakota (2022 Spring) TBD (TBD) 0.00% 0.00% 0.00%
Zhao, Keren (2022 Spring) TBD (TBD) 0.00% 0.00% 0.00%
Zhao, Keren (2022 Spring) TBD (TBD) 0.00% 0.00% 0.00%