Comments for Exploration and Exploitation 3 https://explochallenge.inria.fr New Challenges, ICML 2012 Workshop - Edinburgh Mon, 04 Jun 2012 12:14:09 +0000 hourly 1 http://wordpress.org/?v=3.3.1 Comment on Challenge discussion by Ku-Chun https://explochallenge.inria.fr/challenge-discussion/#comment-123 Ku-Chun Mon, 04 Jun 2012 12:14:09 +0000 http://explochallenge.inria.fr/?page_id=286#comment-123 It is really great that Yahoo is willing to release these datasets. Lacking real world data is a big concern for contextual bandit researchers(at least for me it is). It is really great that Yahoo is willing to release these datasets. Lacking real world data is a big concern for contextual bandit researchers(at least for me it is).

]]> Comment on Challenge discussion by Jeremie https://explochallenge.inria.fr/challenge-discussion/#comment-122 Jeremie Mon, 04 Jun 2012 11:46:01 +0000 http://explochallenge.inria.fr/?page_id=286#comment-122 It'll be done after the workshop on the Yahoo! webscope program It’ll be done after the workshop on the Yahoo! webscope program

]]> Comment on Challenge discussion by Yildiz https://explochallenge.inria.fr/challenge-discussion/#comment-121 Yildiz Mon, 04 Jun 2012 11:10:55 +0000 http://explochallenge.inria.fr/?page_id=286#comment-121 Yeah, thank you very much for the great competition!!! :) Will you later publish the initial dataset, or the dataset from phase 2, so we can continue offline? Would be great ;) thanks again Yeah, thank you very much for the great competition!!! :)

Will you later publish the initial dataset, or the dataset from phase 2, so we can continue offline? Would be great ;)

thanks again

]]> Comment on Challenge discussion by Ku-Chun https://explochallenge.inria.fr/challenge-discussion/#comment-120 Ku-Chun Sun, 03 Jun 2012 11:50:07 +0000 http://explochallenge.inria.fr/?page_id=286#comment-120 Thank you for hosting such an enjoyable competition, had a great time! Thank you for hosting such an enjoyable competition, had a great time!

]]> Comment on Challenge discussion by Jeremie https://explochallenge.inria.fr/challenge-discussion/#comment-119 Jeremie Fri, 01 Jun 2012 16:55:38 +0000 http://explochallenge.inria.fr/?page_id=286#comment-119 Precise time limit is Samoa Standard Time (that means as long as it is June 2th somewhere on earth you can submit). After the deadline you cannot submit, but you can indicate me your favorite old submission. This submission will replace your "best" one. About "paying attention" this is about not exclude a submission because of a trivial problem on Phase 2 datas. Precise time limit is Samoa Standard Time (that means as long as it is June 2th somewhere on earth you can submit).

After the deadline you cannot submit, but you can indicate me your favorite old submission. This submission will replace your “best” one. About “paying attention” this is about not exclude a submission because of a trivial problem on Phase 2 datas.

]]> Comment on Challenge discussion by exploreit https://explochallenge.inria.fr/challenge-discussion/#comment-118 exploreit Fri, 01 Jun 2012 16:46:15 +0000 http://explochallenge.inria.fr/?page_id=286#comment-118 Hi Jeremie, Would we get a chance after June 2 to submit code for Phase 2 evaluation (taking care of feature removal, tuning constants etc) ? or would you just use our best submission file for phase 2 data? I did not understand when you said "I’ll pay attention to get them working on phase 2 data". Also, what time exactly will the last submission be accepted tomorrow? Hi Jeremie,

Would we get a chance after June 2 to submit code for Phase 2 evaluation (taking care of feature removal, tuning constants etc) ? or would you just use our best submission file for phase 2 data? I did not understand when you said “I’ll pay attention to get them working on phase 2 data”.

Also, what time exactly will the last submission be accepted tomorrow?

]]> Comment on Challenge discussion by Jeremie https://explochallenge.inria.fr/challenge-discussion/#comment-117 Jeremie Tue, 29 May 2012 12:13:42 +0000 http://explochallenge.inria.fr/?page_id=286#comment-117 Yes it is :) The second round is to avoid all possible overfitting. Yes it is :)
The second round is to avoid all possible overfitting.

]]> Comment on Challenge discussion by jamh https://explochallenge.inria.fr/challenge-discussion/#comment-116 jamh Tue, 29 May 2012 12:08:18 +0000 http://explochallenge.inria.fr/?page_id=286#comment-116 Hi Dr. exploreit. As I understand, for the first part of the challenge, there is no final round. The winner will be the one with higher score passed the deadline. For the second round I am not clear now how it will be. As I understand there will be a final round with more data as Jeremie explains. Hi Dr. exploreit.

As I understand, for the first part of the challenge, there is no final round. The winner will be the one with higher score passed the deadline.

For the second round I am not clear now how it will be. As I understand there will be a final round with more data as Jeremie explains.

]]> Comment on Challenge discussion by Jeremie https://explochallenge.inria.fr/challenge-discussion/#comment-115 Jeremie Mon, 28 May 2012 21:22:59 +0000 http://explochallenge.inria.fr/?page_id=286#comment-115 The final dataset is 4 times bigger so I expect to have lower variance. Anyway if the scores appears to be close, I'll check with some t-tests The final dataset is 4 times bigger so I expect to have lower variance. Anyway if the scores appears to be close, I’ll check with some t-tests

]]> Comment on Challenge discussion by exploreit https://explochallenge.inria.fr/challenge-discussion/#comment-114 exploreit Mon, 28 May 2012 20:16:40 +0000 http://explochallenge.inria.fr/?page_id=286#comment-114 I am noticing that random seed variations between runs can cause a swing of upto +/-10 points. since there nearly 10 people at the top within a difference of 20, this could be a significant effect. Do you have any suggestions for fixing this issue? Will the best submissions of each participant be run a number of times and the best/average score taken? I am noticing that random seed variations between runs can cause a swing of upto +/-10 points. since there nearly 10 people at the top within a difference of 20, this could be a significant effect. Do you have any suggestions for fixing this issue? Will the best submissions of each participant be run a number of times and the best/average score taken?

]]>