Code for paper "Exploration in Online Advertising Systems with Deep Uncertainty-Aware Learning"
Code for reproducing most of the results in the paper:
Exploration in Online Advertising Systems with Deep Uncertainty-Aware Learning. Chao Du, Zhifeng Gao, Shuo Yuan, Lining Gao, Ziyan Li, Yifan Zeng, Xiaoqiang Zhu, Jian Xu, Kun Gai and Kuang-chih Lee. SIGKDD 2021.
This project is tested under the following environment settings:
http://ml.cs.tsinghua.edu.cn/~chaodu/static/files/DUAL-data.tar.gz
.tar -zxvf DUAL-data.tar.gz
to extract the data files.data
folder should be placed in the base directory of the project.We provide a convenient experiment launcher to produce results using multiple different seeds.
Try python batchrunner_dnn.py
to train the DNN
architecture with DUAL using seeds [0 - 31].
If everything goes well, this should reproduce the result "$0.7755 \pm 0.0020$" in Table 1.
Our code is developed based on the mouna99/dien project.