BriefGPT.xyz
Apr, 2017
从大规模数据中学习分段线性模型，用于广告点击率预测
Learning Piece-wise Linear Models from Large Scale Data for Ad Click Prediction
HTML
PDF
Kun Gai, Xiaoqiang Zhu, Han Li, Kai Liu, Zhe Wang
TL;DR
本文介绍了一个用于大规模稀疏数据的CTR预测的工业级模型LS-PLM，该模型能够在分布式系统中高效捕捉来自海量数据中的非线性模式，并在阿里巴巴在线广告系统中服务于数亿用户。
Abstract
ctr prediction
in real-world business is a difficult
machine learning
problem with
large scale
nonlinear
→