推荐系统概况：传统CTR深度学习CTR GraphEmbedding多任务学习梳理

Posted 2021-04-19 Coggle数据科学

写在前面

本文主要对推荐系统算法模型进行全面梳理，并整理成四张卡片，非常适合快速复习回顾。

推荐系统概况：传统CTR、深度学习CTR、Graph Embedding、多任务学习。

对比内容包括：算法思想、优缺点、网络结构、公式、代码实现等方面。

【推荐系统】专栏历史文章：

一、传统CTR

涵盖：协同过滤、矩阵分解、LR、FM、FFM、GBDT+LR、XGBoost+LR

二、深度学习CTR

涵盖：NeuralCF、Deep Crossing、FNN、PNN、Wide&Deep、Deep&Cross、DeepFM、AFM、DIN、DIEN、DLRM

三、Graph Embedding

涵盖：Word2Vec、Item2Vec、DeepWalk、Node2Vec、EGES

四、多任务学习

涵盖：ESMM、MMoE

参考文献

[1] Rendle, Steffen. "Factorization Machines." 2011.

[2] Mcartney, D . "Proceedings of the Eighth International Workshop on Data Mining for Online Advertising."Eighth International Workshop on Data Mining for Online AdvertisingACM, 2014.

[3] Zhang, Weinan , T. Du , and J. Wang . "Deep Learning over Multi-field Categorical Data: A Case Study on User Response Prediction." (2016).

[4] Product-base Neural Networks for user responses

[5] Xiangnan He, and Tat-Seng Chua. "Neural Factorization Machines for Sparse Predictive Analytics."the 40th International ACM SIGIR ConferenceACM, 2017.

[6] Yang, Yi.et. "Operation-aware Neural Networks for User Response Prediction.".

[7] Juan, Yuchin, Lefortier, Damien, and Chapelle, Olivier. "Field-aware Factorization Machines in a Real-world Online Advertising System.".

[8] Xiao, Jun, Ye, Hao, He, Xiangnan, Zhang, Hanwang, Wu, Fei, & Chua, Tat-Seng. . Attentional factorization machines: learning the weight of feature interactions via attention networks.

[9] Cheng, Heng Tze , et al. "Wide & Deep Learning for Recommender Systems." (2016).

[10] Guo, Huifeng, Tang, Ruiming, Ye, Yunming, Li, Zhenguo, & He, Xiuqiang. . Deepfm: a factorization-machine based neural network for ctr prediction.

[11] Wang, Ruoxi, Fu, Bin, Fu, Gang, & Wang, Mingliang. . Deep & cross network for ad click predictions.

[12] Lian, Jianxun, Zhou, Xiaohuan, Zhang, Fuzheng, Chen, Zhongxia, Xie, Xing, & Sun, Guangzhong. . Xdeepfm: combining explicit and implicit feature interactions for recommender systems.

[13] Song, Weiping, Shi, Chence, Xiao, Zhiping, Duan, Zhijian, Xu, Yewen, & Zhang, Ming et. . Autoint: automatic feature interaction learning via self-attentive neural networks.

高清卡片获取方式

竞赛、学习交流群已成立

一起进行推荐系统、广告算法、数据算法竞赛交流分享

如果加入了之前的社群不需要重复添加！

若进群失败，可在Coggle后台回复【竞赛群】

即可得到最新的二维码！