This repository provides a comprehensive implementation of a deep neural network-based recommendation system similar to YouTube's. The repo is organized to incl

I have surveyed the technology and papers of CTR & Recommender System, and implemented 25 common-used models with Pytorch for reusage. （对工业界学术界的CTR推荐调研并实现25个算法模型，2023）

The source code for our paper "Scenario-Adaptive Feature Interaction for Click-Through Rate Prediction" (accepted by KDD2023 Applied Science Track), which proposes a model for Multi-Scenario/Multi-Domain Recommendation.

Source code for GIFT (CIKM 22)

Click-Through Rate (CTR) prediction is a crucial task in online advertising aimed at estimating the likelihood of a user clicking on an ad.

Get AUC 0.809 at Criteo dataset by MLP

Get AUC 0.794 at Movielens 20M dataset

SmurphCast – percentage‑first time‑series forecasting (churn, CTR, conversion, retention) with additive + GBM + ES‑RNN stacking and automatic model selection. 100 % Python, CPU‑friendly, explainable.

SmurphCast – percentage‑first time‑series forecasting (churn, CTR, conversion, retention) with additive + GBM + ES‑RNN stacking and automatic model selection. 100 % Python, CPU‑friendly, explainable.

Logging some competitions in the field of data mining.

This respository is used to log some deep learning based recommendation models.

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

Performance analysis of DeepFM Recommender System on CTR Dataset

This project analyzes click-through rates (CTR) for advertising campaigns using a dataset of ad impressions and clicks. The goal is to derive insights and improve advertising strategies based on the analysis.

웹 광고 클릭률 예측 AI 경진대회, DACON (2024.05.07 ~ 2024.06.03)

This project analyzes social media ad campaign performance using a dataset of user interactions. It uncovers which platforms, ad categories, and user segments drive the highest engagement, conversions, and simulated revenue.

datawhale&科大讯飞举办的学习挑战赛————“广告点击率预估挑战赛” Rank9 方案

Machine learning project predicting Click-Through Rates using XGBoost with hyperparameter tuning (Test RMSE: 0.059)