-
https://www.cs.cornell.edu/people/tj/publications/yue_joachims_09a.pdf
-
I have been playing around with your code and reading your paper, and I wish to apply it to a personal problem with picking stocks to invest. I was studying what a preference matrix is and the example…
-
## ざっくり言うと
- ランク学習をオンラインで行う場合にパラメータ更新の勾配の分散を小さく抑えるという内容
- オンライン学習は"dueling bandit"という，ランダムなアップデートによりモデルが改善するかどうかを判断しながら学習していく方法を用いる
- 1回のアップデート時に用いた文書が張る空間にアップデートの勾配を制限することで，分散を小さく抑える(Document Spac…
-
On the NEXT frontend GUI, Dashboard > Select project > new experiment > \* presents a bunch of options which are not clearly explained. At first glance, it's not clear what "context type" means or wha…
-
< Placeholder >
timeline: April 2023 - April 2027.
[Key historical 2016 issue of thesis topic](https://github.com/Tribler/tribler/issues/2250)
ToDo: 6 weeks hands-on Python onboarding project. …
-
For whatever reason, whenever I click edit or share on my 46-page campaign guide I made for my players, I cannot see any of the guide. It's a single, grey page with "1/46" written in the top left corn…
-
It is missing a prototype and most algs have not been ported over. The algs that are there also do not seem correct
-
While the launch script examples work its not obvious how to change the behavior even a little without digging into the code.
For example, in dueling bandits its not obvious that to not have a "cont…
-
Experiment design interface says "oops there is an error" if we try to use the "random sampling" algorithm instead of the default algorithm (Dueling Bandits).
-
"server error" if you try to download the data paricipant response data (Dueling Bandits)