Optimization Method
Online Optimization，Parallel SGD，FTRL等优化方法，实用并且能够给出直观解释的文章
- Google Vizier A Service for Black-Box Optimization.pdf
Google的深度学习自动调参框架Vizier - 在线最优化求解(Online Optimization)-冯扬.pdf
非常推荐冯扬的这个教程，把在线优化问题讲的非常透 - Hogwild A Lock-Free Approach to Parallelizing Stochastic Gradient Descent.pdf
- Parallelized Stochastic Gradient Descent.pdf
- A Survey on Algorithms of the Regularized Convex Optimization Problem.pptx
- Follow-the-Regularized-Leader and Mirror Descent- Equivalence Theorems and L1 Regularization.pdf
- A Review of Bayesian Optimization.pdf
- Taking the Human Out of the Loop- A Review of Bayesian Optimization.pdf
- 非线性规划.doc
CTR Prediction
作为计算广告的核心，CTR预估永远是研究的热点，下面每一篇都是非常流行的文章，推荐逐一精读
- Deep Crossing- Web-Scale Modeling without Manually Crafted Combinatorial Features.pdf
- Learning Piece-wise Linear Models from Large Scale Data for Ad Click Prediction.pdf
阿里提出的Large Scale Piece-wise Linear Model (LS-PLM) CTR预估模型 - [GBDT+LR]Practical Lessons from Predicting Clicks on Ads at Facebook.pdf
- [FNN]Deep Learning over Multi-field Categorical Data.pdf
- Entire Space Multi-Task Model_ An Effective Approach for Estimating Post-Click Conversion Rate.pdf
- Deep Interest Network for Click-Through Rate Prediction.pdf
- Bid-aware Gradient Descent for Unbiased Learning with Censored Data in Display Advertising.pdf
RTB 中训练 CTR 模型数据集是赢得出价的广告，预测时的样本却是所有候选的广告，也就是训练集和测试集的分布不一致，这篇文章就是要消除这样的 bias - [Multi-Task]An Overview of Multi-Task Learning in Deep Neural Networks.pdf
- Ad Click Prediction a View from the Trenches.pdf
Google大名鼎鼎的用FTRL解决CTR在线预估的工程文章，非常经典。 - [PNN]Product-based Neural Networks for User Response Prediction.pdf
- Image Matters- Visually modeling user behaviors using Advanced Model Server.pdf
阿里提出引入商品图像特征的（Deep Image CTR Model）CTR预估模型，并介绍其分布式机器学习框架 Advanced Model Server (AMS) - [Wide & Deep]Wide & Deep Learning for Recommender Systems.pdf
- [DeepFM]- A Factorization-Machine based Neural Network for CTR Prediction.pdf
- Logistic Regression in Rare Events Data.pdf
样本稀少情况下的LR模型训练，讲的比较细 - Deep & Cross Network for Ad Click Predictions.pdf
Google 在17年发表的 Deep&Cross 网络，类似于 Wide&Deep, 比起 PNN 只做了特征二阶交叉，Deep&Cross 理论上能够做任意高阶的特征交叉 - Learning Deep Structured Semantic Models for Web Search using Clickthrough Data.pdf
- Adaptive Targeting for Online Advertisement.pdf
一篇比较简单但是全面的CTR预估的文章，有一定实用性
papers 收集
在线课程
- Introduction to Computational Advertising - Stanford
- 计算广告学 - 刘鹏
- 计算广告2.0 - 刘鹏
- 计算广告学概论 - 百度
- 计算广告学之搜索引擎广告原理 -百度
- 计算广告学之内容匹配广告&展示广告原理、技术和实践 - 百度
- 计算广告学-2014 - 百度
书籍
系列
案例
协议标准
参考文献
[1] https://www.jianshu.com/p/8c591feb9fc4
[2] https://github.com/duboya/CTR-Prediction
扫一扫
2万+
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
