-
-
(mainly a reference, and possibly a reference for unit tests)
https://cran.r-project.org/web/packages/drgee/index.html
http://www.degruyter.com/view/j/em.2015.4.issue-1/em-2014-0021/em-2014-0021.xml…
-
## 論文リンク
https://arxiv.org/abs/1906.02443
## 公開日(yyyy/mm/dd)
2019/07
## 概要
encoder への入力と decode への入力の両方で adversarial training を適用して機械翻訳の性能を高めたという論文。
単語置き換えをベースに adversarial training のための adv…
-
@kbattocchi
Hi Keith,
How would you recommend handling a case where we know the true treatment probabilities? I'd prefer to use them to avoid having to fit the model_propensity (in a doubly rob…
-
When applying the "average_treatment_effect" function to the output of a "multi_arm_causal_forest" with K>2 arms, a table with K-1 ATEs is returned along with standard error estimates for those ATEs. …
-
Thank you for this useful repo! I have a question, lets say there is logged data you want to use to train and evaluate a new policy. The logged data is something like where the context features descr…
-
As stated in the title, if the propensity model in doubly robust learning is nearly perfect, then the inverse propensity weight would be extremely high, causing Y^{DR}_{i,t} to be inaccurate and the r…
-
## 論文リンク
https://arxiv.org/pdf/1906.02443.pdf
## 概要
NMTにおいて損失を利用したwhite-boxでの敵対的サンプル作成法を提案
作成した敵対的サンプルを用いた学習によるロバストなNMTモデルを提案
-
## 論文URL
https://usaito.github.io/files/RecSys2020_DRMetric.pdf
## 著者
YUTA SAITO
## 会議
RecSys 20
## 資料
https://vimeo.com/455947055
## 背景
post-clickのフィードバックに関して考えるとき、従来の推定量では…
-
Currently, working on a DML meta-estimator that we can plug in any arbitrary trees. This could probably replace the DML set of classes which are pretty confusing (to me).
Another set of API we wou…