Please wait a minute...
 首页  期刊介绍 期刊订阅 联系我们 横山亮次奖 百年刊庆
 
最新录用  |  预出版  |  当期目录  |  过刊浏览  |  阅读排行  |  下载排行  |  引用排行  |  横山亮次奖  |  百年刊庆
清华大学学报(自然科学版)  2018, Vol. 58 Issue (4): 374-379    DOI: 10.16511/j.cnki.qhdxxb.2018.26.021
  计算机科学与技术 本期目录 | 过刊浏览 | 高级检索 |
基于特征关联模型的广告点击率预测
沈方瑶1, 戴国骏1, 代成雷2, 郭鸿杰3, 张桦1
1. 杭州电子科技大学 计算机学院, 杭州 310018;
2. 浙江大学 数学科学学院, 杭州 310058;
3. 哈尔滨工业大学 计算机科学与技术学院, 哈尔滨 150001
CTR prediction for online advertising based on a features conjunction model
SHEN Fangyao1, DAI Guojun1, DAI Chenglei2, GUO Hongjie3, ZHANG Hua1
1. School of Computer Science and Technology, Hangzhou Dianzi University, Hangzhou 310018, China;
2. Department of Mathematical Science, Zhejiang University, Hangzhou 310058, China;
3. School of Computer Science and Technology, Harbin Institute of Technology, Harbin 150001, China
全文: PDF(1031 KB)  
输出: BibTeX | EndNote (RIS)      
摘要 点击率(click-through rate,CTR)预测是互联网公司中重要的研究课题,预测结果与上下文、用户属性和广告属性息息相关,CTR的有效预测对提高广告公司的收入至关重要。该文在对传统逻辑回归(logistic regression,LR)模型的相关原理和参数优化算法介绍的基础上,抽离出用户特征和广告特征,将用户与广告之间特征的关联信息添加到Sigmoid函数中得到一种特征关联模型。与以往求解方法不同,该方法采用在线最优化算法FTRL(follow-the-regularized-leader)提高参数计算效率,采用混合正则化来防止训练过拟合。真实的广告数据集上的实验结果表明:该方法与传统的模型和方法相比具有更好的预测精度、效率、参数敏感性和可靠性。
服务
把本文推荐给朋友
加入引用管理器
E-mail Alert
RSS
作者相关文章
沈方瑶
戴国骏
代成雷
郭鸿杰
张桦
关键词 点击率预测特征关联在线最优化混合正则项    
Abstract:Click-through rate (CTR) predictions are important for internet companies. The CTR is closely related to the context, user attributes and advertising attributes, with effective CTR predictions essential for improving company revenue. The traditional LR model was optimized to predict the relationship between the user and advertiser characteristics for the CTR which were added to the Sigmoid function to obtain a new features conjunction model. The online optimization algorithm follow-the-regularized-leader (FTRL) was used to improve the efficiency of the parameter, and the mixed regularization was used to prevent over fitting. Tests on a real-world advertising dataset show that this method has better accuracy, efficiency, parameter sensitivity and reliability compared with previous algorithms.
Key wordsclick-through rate (CTR)    features conjunction    online optimization    mixed regularization
收稿日期: 2017-12-27      出版日期: 2018-04-15
ZTFLH:  TP391.1  
基金资助:国家自然科学基金联合基金项目(U1509216);国家自然科学基金资助项目(61471150)
通讯作者: 张桦,副教授,E-mail:zhangh@hdu.edu.cn     E-mail: zhangh@hdu.edu.cn
作者简介: 沈方瑶(1992-),女,博士研究生。
引用本文:   
沈方瑶, 戴国骏, 代成雷, 郭鸿杰, 张桦. 基于特征关联模型的广告点击率预测[J]. 清华大学学报(自然科学版), 2018, 58(4): 374-379.
SHEN Fangyao, DAI Guojun, DAI Chenglei, GUO Hongjie, ZHANG Hua. CTR prediction for online advertising based on a features conjunction model. Journal of Tsinghua University(Science and Technology), 2018, 58(4): 374-379.
链接本文:  
http://jst.tsinghuajournals.com/CN/10.16511/j.cnki.qhdxxb.2018.26.021  或          http://jst.tsinghuajournals.com/CN/Y2018/V58/I4/374
  图1 算法1
  表1 本文提出的方法与其他方法性能对比
  图2 LogGloss和<em>k</em>的关系
  图3 本文提出方法的收敛性
  图4 不同数据量对 LogGloss的影响
[1] CHAPELLE O, MANAVOGLU E, ROSALES R. Simple and scalable response prediction for display advertising[J]. ACM Transactions on Intelligent Systems and Technology, 2015, 5(4):1-34.
[2] AGARWAL A, CHAPELLE O, DUDÍK M, et al. A reliable effective terascale linear learning system[J]. The Journal of Machine Learning Research, 2014, 15(1):1111-1133.
[3] 黄璐, 林川杰, 何军, 等. 融合主题模型和协同过滤的多样化移动应用推荐[J]. 软件学报, 2017, 28(3):708-720. HUANG L, LIN C J, HE J, et al. Diversified mobile app recommendation combining topic model and collaborative filtering[J]. Journal of Software, 2017, 28(3):708-720. (in Chinese)
[4] GRAEPEL T, CANDELA J Q, BORCHERT T, et al. Web-scale Bayesian click-through rate prediction for sponsored search advertising in microsoft's bing search engine[C]//Proceedings of the 27th International Conference on Machine Learning. Haifa, Israel:ACM, 2010:13-20.
[5] MA J, SAUL L K, SAVAGE S, et al. Learning to detect malicious URLs[J]. ACM Transactions on Intelligent Systems and Technology, 2011, 2(3):1-24.
[6] BAI Y Q, SHEN K J. Alternating direction method of multipliers for L1-L2-regularized logistic regression model[J]. Journal of the Operations Research Society of China, 2016, 4(2):243-253.
[7] QUAN D Y, YIN L H, GUO Y C. Assessing the disclosure of user profile in mobile-aware services[C]//Proceedings of the 11th International Conference on Information Security and Cryptology. Beijing, China:Springer, 2015:451-467.
[8] ZINKEVICH M. Online convex programming and generalized infinitesimal gradient ascent[C]//Proceedings of the 20th International Conference on Machine Learning. Washington, USA:AAIA, 2003:928-936.
[9] LANGFORD J, LI L H, ZHANG T. Sparse online learning via truncated gradient[J]. The Journal of Machine Learning Research, 2009, 10:777-801.
[10] DUCHI J, SINGER Y. Efficient online and batch learning using forward backward splitting[J]. The Journal of Machine Learning Research, 2009, 10:2899-2934.
[11] LIN X. Dual averaging methods for regularized stochastic learning and online optimization[J]. The Journal of Machine Learning Research, 2010, 11:2543-2596.
[12] MCMAHAN H B, HOLT G, SCULLEY D, et al. Ad click prediction:A view from the trenches[C]//Proceedings of the 19th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. Chicago, USA:ACM, 2013:1222-1230.
[13] MCMAHAN H B. Follow-the-regularized-leader and mirror descent:Equivalence theorems and L1-regularization[C]//Proceedings of the 14th International Conference on Artificial Intelligence and Statistics. Fort Lauderdale, USA:JMLR, 2011:525-533.
[14] BILGIC B, CHATNUNTAWECH I, FAN A P, et al. Fast image reconstruction with L2-regularization[J]. Journal of Magnetic Resonance Imaging, 2014, 40(1):181-191.
[15] TIBSHIRANI R. Regression shrinkage and selection via the Lasso[J]. Journal of the Royal Statistical Society, 1996, 58(1):267-288.
[16] YAN L, LI W J, XUE R G, et al. Coupled group lasso for web-scale CTR prediction in display advertising[C]//Proceedings of the 31st International Conference on Machine Learning. Beijing, China:ACM, 2014:802-810.
[1] 张婧, 黄德根, 黄锴宇, 刘壮, 孟祥主. 基于λ-主动学习方法的中文微博分词[J]. 清华大学学报(自然科学版), 2018, 58(3): 260-265.
[2] 张仰森, 郑佳, 黄改娟, 蒋玉茹. 基于双重注意力模型的微博情感分析方法[J]. 清华大学学报(自然科学版), 2018, 58(2): 122-130.
[3] 哈妮克孜·伊拉洪, 古力米热·依玛木, 玛依努尔·阿吾力提甫, 姑丽加玛丽·麦麦提艾力, 艾斯卡尔·艾木都拉. 维吾尔语感叹句语调起伏度[J]. 清华大学学报(自然科学版), 2017, 57(12): 1254-1258.
[4] 王素格, 李大宇, 李旸. 基于联合模型的商品口碑数据情感挖掘[J]. 清华大学学报(自然科学版), 2017, 57(9): 926-931.
[5] 郭少茹, 张虎, 钱揖丽, 李茹, 杨陟卓, 顾兆军, 马淑晖. 面向高考阅读理解的句子语义相关度[J]. 清华大学学报(自然科学版), 2017, 57(6): 575-579,585.
[6] 米吉提·阿不里米提, 艾克白尔·帕塔尔, 艾斯卡尔·艾木都拉. 基于层次化结构的语言模型单元集优化[J]. 清华大学学报(自然科学版), 2017, 57(3): 257-263.
[7] 赛牙热·依马木, 热依莱木·帕尔哈提, 艾斯卡尔·艾木都拉, 李志军. 基于不同关键词提取算法的维吾尔文本情感辨识[J]. 清华大学学报(自然科学版), 2017, 57(3): 270-273.
[8] 热合木·马合木提, 于斯音·于苏普, 张家俊, 宗成庆, 艾斯卡尔·艾木都拉. 基于模糊匹配与音字转换的维吾尔语人名识别[J]. 清华大学学报(自然科学版), 2017, 57(2): 188-196.
[9] 阿不都萨拉木·达吾提, 于斯音·于苏普, 艾斯卡尔·艾木都拉. 类别区分词与情感词典相结合的维吾尔文句子情感分类[J]. 清华大学学报(自然科学版), 2017, 57(2): 197-201.
[10] 高莹莹, 朱维彬. 面向情感语音合成的言语情感描述与预测[J]. 清华大学学报(自然科学版), 2017, 57(2): 202-207.
[11] 刘泽文, 丁冬, 李春文. 基于条件随机场的中文短文本分词方法[J]. 清华大学学报(自然科学版), 2015, 55(8): 906-910,915.
[12] 王丙坤, 黄永峰, 李星. 基于多粒度计算和多准则融合的情感分类[J]. 清华大学学报(自然科学版), 2015, 55(5): 497-502.
[13] 陈元琳, 柴跃廷, 刘义, 徐扬. 基于群体偏好的交易评价可信度[J]. 清华大学学报(自然科学版), 2015, 55(5): 558-564,571.
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
版权所有 © 《清华大学学报(自然科学版)》编辑部
本系统由北京玛格泰克科技发展有限公司设计开发 技术支持:support@magtech.com.cn