Dear Zindians,
Well done to all who participated in this challenge!
At Zindi, we pride ourselves on providing clients with solutions that are useful and will provide impact. This means that we review your submitted code to make sure that solutions comply with the competition rules and resource restrictions.
For this particular competition, an additional requirement was made in the middle of the competition: specifically, we required participants not to use the damage feature in their solution. This requirement is in line with Zindi's aim to provide solutions of value to the client.
Zindi is a place for competitive data science, but more so, it is a place to develop real life business and impact skills. We expect our participants to their skill and understanding to provide competition hosts with a valuable solution.
Based on the above requirements and through code review, users who did not submit code for review or users who used the damage feature in their solution were disqualified from this competition.
This is inline with the competition rules:
The leaderboard is now sealed, and you can download your participation certificate next to your username on the leaderboard link for the competition.
Regards,
Amy
Dear Zindi Team,
I can't find a item related to the damage feature in the competition rules.
I consider disqualifications for this reason is unjustified.
The leakage of private labels poses a risk, as individuals are aware that the value range of certain images is set at 0. Consequently, individuals may manipulate the range of specific images to deceive private scoring. Exercise caution when employing postprocessing solutions to boost scores.