AI Rank
EN
← 开源
qiaoguan

deep-ctr-prediction

CTR prediction models based on deep learning(基于深度学习的广告推荐CTR预估模型)

Model DevelopmentClassical MLPython
在 GitHub 打开
增长势头
+024 小时新增 Star0.0%
938
Star
277
Fork
+0
本周
1
贡献者
创建于 2019-02-17 · 更新于 2026-08-27 · 今日第 12517 名
主要开发者
qiaoguan

qiaoguan/deep-ctr-prediction 排名第 12,517，有 938 个星标。

同类别比较

qiaoguan/deep-ctr-prediction 排名第 12,517，此处有 8 个同类条目；mst272/LLM-Dojo 排名第 12,515。

同类别比较
名称排名指标
nv-tlabs/GSCNN12,510
语言:
Python
星标数:
939
分叉数:
198
简介:
Gated-Shape CNN for Semantic Segmentation (ICCV 2019)
THUDM/P-tuning12,512
语言:
Python
星标数:
938
分叉数:
112
简介:
A novel method to tune language models. Codes and datasets for paper ``GPT understands, too''.
esa/pagmo212,513
语言:
C++
星标数:
938
分叉数:
190
简介:
A C++ platform to perform parallel computations of optimisation tasks (global and local) via the asynchronous generalized island model.
mst272/LLM-Dojo12,515
语言:
Python
星标数:
938
分叉数:
85
简介:
轻量级 LLM Post-training 框架，支持 SFT、RLVR、On-Policy KD、Guide KD 及混合训练；实现单轮/多轮 Guide 蒸馏、多教师蒸馏、Reward 混合训练与自动化数据分流👩‍🎓👨‍🎓
Farama-Foundation/Shimmy12,519
语言:
Python
星标数:
937
分叉数:
32
简介:
PettingZoo and Gymnasium bindings for popular reinforcement learning environments outside of Farama
eaplatanios/tensorflow_scala12,521
语言:
Scala
星标数:
937
分叉数:
89
简介:
TensorFlow API for the Scala Programming Language
mnicnc404/CartoonGan-tensorflow12,524
语言:
Python
星标数:
937
分叉数:
238
简介:
Generate your own cartoon-style images with CartoonGAN (CVPR 2018), powered by TensorFlow 2.0 Alpha.
acbull/pyHGT12,529
语言:
Python
星标数:
936
分叉数:
173
简介:
Code for "Heterogeneous Graph Transformer" (WWW'20), which is based on pytorch_geometric
关键参数与规格
关键参数与规格
参数值
语言Python
类别Model Development
子类别Classical ML
星标数938
分叉数277
贡献者1
创建日期2019-02-17
更新日期2026-08-27
1 天星标数0
7 天星标数0
排名12,517
上次排名12,441
排名变化-76
7 天星标排名13,468
上次 7 天星标排名13,478
7 天星标排名变化10
总星标排名11,703
上次总星标排名11,686
总星标排名变化-17
国家China
README 摘录

deep-ctr-prediction 一些广告算法(CTR预估)相关的DNN模型 wide&deep 可以参考official/wide_deep deep&cross deepfm ESMM Deep Interest Network ResNet xDeepFM AFM(Attentional FM) Transformer FiBiNET 代码使用tf. estimator构建, 数据存储为tfrecord格式(字典，key:value), 采用tf. 特征工程定义在input_fn,模型定义在model_fn

,实现特征和模型代码分离,特征工程代码只用修改input_fn,模型代码只用修改model_fn。 数据默认都是存在hadoop，可以根据自己需求存在本地, 特征工程和数据的处理可以参考Google开源的wide&deep模型(不使用tfrecord

格式, 代码在official/wide_deep) All codes are written based on tf. 10 参考文献 【1】Heng-Tze Cheng, Levent Koc et all.

常见问题

记录了多少星标和分叉？

qiaoguan/deep-ctr-prediction 有 938 个星标和 277 个分叉。

仓库排名是多少？

qiaoguan/deep-ctr-prediction 排名第 12,517。

记录的语言和类别是什么？

qiaoguan/deep-ctr-prediction 有 938 个星标，语言为 Python，类别为 Model Development。

哪个仓库排名最接近？

qiaoguan/deep-ctr-prediction 与 mst272/LLM-Dojo 比较，后者排名第 12,515，有 938 个星标。

README

deep-ctr-prediction

一些广告算法(CTR预估)相关的DNN模型

  • wide&deep 可以参考official/wide_deep

  • deep&cross

  • deepfm

  • ESMM

  • Deep Interest Network

  • ResNet

  • xDeepFM

  • AFM(Attentional FM)

  • Transformer

  • FiBiNET

代码使用tf.estimator构建, 数据存储为tfrecord格式(字典，key:value), 采用tf.Dataset API, 加快IO速度，支持工业级的应用。特征工程定义在input_fn,模型定义在model_fn,实现特征和模型代码分离,特征工程代码只用修改input_fn,模型代码只用修改model_fn。数据默认都是存在hadoop，可以根据自己需求存在本地, 特征工程和数据的处理可以参考Google开源的wide&deep模型(不使用tfrecord格式, 代码在official/wide_deep)

All codes are written based on tf.estimator API, the data is stored in tfrecord format(dictionary, key:value), and the tf.Dataset API is used to speed up IO speed, it support industrial applications. Feature engineering is defined in input_fn, model function is defined in model_fn, the related code of feature engineering and model function is completely separated, the data is stored in hadoop by default, and can be locally stored according to your own need. The feature engineering and data processing can refer to Google's open source wide&deep model(without tfrecord format, codes are available at official/wide_deep)

Requirements

  • Tensorflow 1.10

参考文献

【1】Heng-Tze Cheng, Levent Koc et all. "Wide & Deep Learning for Recommender Systems," In 1st Workshop on Deep Learning for Recommender Systems,2016.

【2】Huifeng Guo et all. "DeepFM: A Factorization-Machine based Neural Network for CTR Prediction," In IJCAI,2017.

【3】Ruoxi Wang et all. "Deep & Cross Network for Ad Click Predictions," In ADKDD,2017.

【4】Xiao Ma et all. "Entire Space Multi-Task Model: An Effective Approach for Estimating Post-Click Conversion Rate," In SIGIR,2018.

【5】Guorui Zhou et all. "Deep Interest Network for Click-Through Rate Prediction," In KDD,2018.

【6】Kaiming He et all. "Deep Residual Learning for Image Recognition," In CVPR,2016.

【7】Jianxun Lian et all. "xDeepFM: Combining Explicit and Implicit Feature Interactions for Recommender Systems," In KDD,2018.

【8】Jun Xiao et all. "Attentional Factorization Machines: Learning the Weight of Feature Interactions via Attention Networks," In IJCAI, 2017.

【9】Ashish Vasmani et all. "Attention is All You Need," In NIPS, 2017.

【10】Tongwen et all. "FiBiNET: Combining Feature Importance and Bilinear feature Interaction for Click-Through Rate Prediction," In RecSys, 2019.

排名相邻
同类项目