I am a participant in the above challenge. But I have a problem/question:
In the second public data set, "ARCENE", you have mentioned that it is a Binary Classification problem. But the target variable is multi-class with hundreds of classes. Is this an error or do you want me to change it into 1 or 0 arbitrarily?
Please clarify.
Thanks,
Hi,
Arcene is indeed used as a binary classification problem for this challenge, you can download the data here:
https://competitions.codalab.org/my/datasets/download/2d962a53-5eb9-46dd-b241-70ce10988414
Best
Hi:
I might be missing something , if so please clarify for the benefit of others as well.
I am looking at arcene_train.data and the first variable is shown as target variable in .info file.
So I am displaying the first few rows of arcene's target variable here:
0 0
1 0
2 0
3 0
4 105
5 38
6 76
7 47
8 0
9 38
10 0
11 17
Does this look like a binary_class target variable? Or am I missing something? Please let us know where or what is the target variable in ARCENE?
Thanks.
Ram
The target variables have extension ".solution"
Posted by: hugo.jair @ March 20, 2018, 10:09 p.m.So if I understand it right, the labels for the train data set are in “.solution” file, correct?
Thanks for being so quick in clarifying.
Ram
What about the labels for the “.valid” set? Are they also there? Thanks
Ram
Please read the instructions at CodaLab or here:
https://www.4paradigm.com/competition/pakdd2018
No labels are provided for validation or testing
Posted by: hugo.jair @ March 20, 2018, 10:40 p.m.