Feature Generation by Convolutional Neural Network for Click-Through Rate Prediction

9 Apr 2019Bin LiuRuiming TangYingzhi ChenJinkai YuHuifeng GuoYuzhou Zhang

Easy-to-use,Modular and Extendible package of deep-learning based CTR models.DeepFM,DeepInterestNetwork(DIN),DeepInterestEvolutionNetwork(DIEN),DeepCrossNetwork(DCN),AttentionalFactorizationMachine(AFM),Neural Factorization Machine(NFM),AutoInt,Deep Session Interest Network(DSIN)..

PDF Abstract
TASK DATASET MODEL METRIC NAME METRIC VALUE GLOBAL RANK RESULT LEADERBOARD
Click-Through Rate Prediction Avazu FGCNN+IPNN AUC 0.7883 # 3
LogLoss 0.3746 # 4
Click-Through Rate Prediction Criteo FGCNN+IPNN AUC 0.8022 # 5
Log Loss 0.5388 # 13
Click-Through Rate Prediction Huawei App Store FGCNN+IPNN AUC 0.9407 # 1
Log Loss 0.1134 # 1