{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/learning-piece-wise-linear-models-from-large","title":"Learning Piece-wise Linear Models from Large Scale Data for Ad Click Prediction","arxiv_id":"1704.05194","date":"2017-04-18","proceeding":null,"authors":["Kun Gai","Xiaoqiang Zhu","Han Li","Kai Liu","Zhe Wang"],"abstract":"CTR prediction in real-world business is a difficult machine learning problem\nwith large scale nonlinear sparse data. In this paper, we introduce an\nindustrial strength solution with model named Large Scale Piece-wise Linear\nModel (LS-PLM). We formulate the learning problem with $L_1$ and $L_{2,1}$\nregularizers, leading to a non-convex and non-smooth optimization problem.\nThen, we propose a novel algorithm to solve it efficiently, based on\ndirectional derivatives and quasi-Newton method. In addition, we design a\ndistributed system which can run on hundreds of machines parallel and provides\nus with the industrial scalability. LS-PLM model can capture nonlinear patterns\nfrom massive sparse data, saving us from heavy feature engineering jobs. Since\n2012, LS-PLM has become the main CTR prediction model in Alibaba's online\ndisplay advertising system, serving hundreds of millions users every day.","url_abs":"http://arxiv.org/abs/1704.05194v1","url_pdf":"http://arxiv.org/pdf/1704.05194v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"learning-piece-wise-linear-models-from-large","repo_url":"https://github.com/UlionTse/mlgb","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"unanswered"}},{"paper_slug":"learning-piece-wise-linear-models-from-large","repo_url":"https://github.com/shenweichen/DeepCTR","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":null},{"paper_slug":"learning-piece-wise-linear-models-from-large","repo_url":"https://github.com/shenweichen/DeepCTR-Torch","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"Apache-2.0"}}],"tasks":[{"task_slug":"click-through-rate-prediction","task_name":"Click-Through Rate Prediction"},{"task_slug":"feature-engineering","task_name":"Feature Engineering"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1704.05194","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}