SEDCN: An improved Deep & Cross Network Recommendation Algorithm based on SENET
Jiao Li, Nanchang Cheng · 2022
Advertising ranking is essential to many Internet companies such as Facebook and Sina Weibo. Among many real-world advertising ranking systems, click-through rate (CTR) prediction plays a central role. There are many proposed models in this field such as logistic regression, factorization machine based models,and deep learning based CTR models. However, many current works calculate the feature interactions in a simple way such as Hadamard product and inner product and they care less about the importance of features. In this paper, we propose a new feature importance network model SEDCN based on deep crossover network. SEDCN can learn the importance of features dynamically through the SENET mechanism. We compare the performance of SEDCN with other depth models, such as DeepFM and DCN.