Reproduce “DeepFM: An End-to-End Wide & Deep Learning Framework for CTR Prediction” on OpenResearch

Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.

Universal app for Apple silicon and Intel. Requires macOS 11 or later.

About this paper

Huifeng Guo, Ruiming Tang, Yunming Ye, Zhenguo Li, Xiuqiang He, Zhenhua Dong

DeepFM predicts click probabilities from sparse user, item, and context features. It combines an FM component for individual-feature effects and pairwise interactions with a neural network for richer combinations; both share learned embeddings and train jointly. In a seven-day Huawei App Market test, DeepFM-D improved CTR by more than 10% over logistic regression, specific to that deployment and comparator.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.