Coupled Group Lasso for Web-Scale CTR Prediction in Display Advertising

Authors: Ling Yan, Wu-Jun Li, Gui-Rong Xue, Dingyi Han

ICML 2014 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experimental results on real-world data sets show that our CGL model can achieve state-of-the-art performance on webscale CTR prediction tasks.
Researcher Affiliation Collaboration Ling Yan YLING0718@SJTU.EDU.CN Shanghai Key Laboratory of Scalable Computing and Systems, Department of Computer Science and Engineering, Shanghai Jiao Tong University, China Wu-Jun Li LIWUJUN@NJU.EDU.CN National Key Laboratory for Novel Software Technology, Department of Computer Science and Technology, Nanjing University, China Gui-Rong Xue GRXUE@ALIBABA-INC.COM Alibaba Group, China Dingyi Han DINGYI.HAN@ALIBABA-INC.COM Alibaba Group, China
Pseudocode Yes Algorithm 1 Alternate Learning for CGL
Open Source Code No The paper does not include an explicit statement about releasing source code for the described methodology, nor does it provide a link to a code repository.
Open Datasets No We conduct our experiment on three real-world data sets connected from Taobao of Alibaba group. We build our data sets from the logs of the advertisements displayed in http://www.taobao.com, one of the most famous C2C e-commerce web sites in China.
Dataset Splits Yes We sample 20% of each training set for validation to specify the hyper-parameters of our CGL model and other baselines.
Hardware Specification Yes We have an MPI-cluster with hundreds of nodes, each of which is a 24-core server with 2.2GHz Intel(R) Xeon(R) E5-2430 processor and 96GB of RAM.
Software Dependencies No The paper mentions using MPI and L-BFGS algorithms but does not provide specific version numbers for any software dependencies or libraries.
Experiment Setup Yes The k in CGL is fixed to 50 in our experiment unless otherwise stated. We vary the values of the hyper-parameter λ and draw the influence on the performance in Figure 3 (b).