This is the second challenge for extracting data from old crossplots. This challenge is different from the previous axes challenge: the scoring metric is Levenshtein distance to support the text theme, and the duration of the challenge has been extended to 6 weeks.
We are looking forward to seeing your ideas about how to solve this deceptively simple challenge.
Good luck out there!
Where is the competition data? Are they the same as the “Identify and Extract Axes Values” competition? They are not available there either.
same question, I also didn’t see the data.
The data for “Detect Text on Crossplots” is the same theme as the recently closed “Identify and Extract Axes Values.” However, the data is unique for each challenge.
Please try to log out and then log back into Xeek. The data only appears for users that are logged in. Note, the “Identify and Extract Axes Values” is no longer available as the challenge has closed.
I log out and log again
There is no “Data” link in “Identify and Extract Axes Values”
I don´t remember but I think I was registered in that competition.
For “Slick in a Haystack - a Localization Challenge” and " Detect Text on Crossplots" I see the Data link
Currently, when a challenge is closed, like “Identify and Extract Axes Values”, we also close data access.
Would you be interested in working on the data for a personal project?
No. What I want is download the data of this competition; as said two messages above, I undestand tha I must use the same data from the “Identify and Extract Axes Values” competition. So to download that data I need access and the competition is closed. Or do I need another way to access the data for this competition " Detect Text on Crossplots"
There might be an issue with a previous login. Please try to log into Xeek using an Incognito (Chrome) or Private (Edge) window. Are you challenged to enter an MFA code? If so, you should be able to navigate to “Detect Text on Crossplots” and check if you have joined the challenge (purple button on the right side of the Overview Tab). If you are signed in and have joined the challenge, you should be able to see the data. If those options don’t work, please email email@example.com, and our Support Team will get you the data and manually connect to the challenge.
Not able to see leader board
It is so weird that no data (except the sample submission and the simple notebook) are there.
The host has to publish the data and extend the deadline
tien2020.le2020 you may not be able to see the data because you haven’t officially joined the challenge. Could you please check your status with the challenge and get back to us on whether you can see the data?
Oh you are right! I see the 2 other files now.
@tien2020.le2020 glad to hear. Good luck and we look forward to seeing you on the Leaderboard soon!
Can we use some state-of-the-art models or do we need to create our models from scratch?
You may use whatever model you wish. However, if you are in consideration of a prize, you must submit your model, including any supplemental materials. Make sure that whatever model you submit has an open or permissive license.
The competition will end soon and I have questions about this:
- In the rules of the competition, I found two conflicting dates for the end of the competition
November 2, 2022 (in the “COMPETITION TIMELINE” part)
November 3rd, 2022 at 23:00 UTC (in the introduction part)
Which one is correct - when is the end?
The judges will score the top 10 submissions on accuracy and interpretability. The accuracy of the submission counts for 90% of the final score. Accuracy will be determined using the same scoring algorithm described above for the Predictive Leaderboard. The top 20% of scores will receive maximum points (90).
Maximum points (90) will be received by the top 20% of the 10 selected (top-2), or 20% of all participants?
- Do the top 10 have time to prepare their notebooks? When is the deadline - also November 2 or 3?
The interpretability metric counts for 10% of the final score. This qualitative metric focuses on the degree of documentation, clearly stating variables for models and using doc strings and markdown. Submissions with exceptional interpretability will receive maximum points (10).
If I understand correctly, then a few maximally documented notebooks can receive a maximum score. If the same participants were in the top 20% from the question 2 - will there be several winners?
- Can you tell us more about the assessment of notebooks? With the same degree of documentation - will a simpler solution be preferred?
- If I trained neural network - should I provide the training code? Will it evaluate too? How well should it be documented?
To add to the previous great questions: Can we submit a repository rather than a single notebook?
@team The competition is almost over, will you answer me?)
hi team, just making sure if my final submission is already received by the team, because I already sent it over email but got no response yet. I hope it’s not caught in your spam filter.