Towards Hybrid-grained Feature Interaction Selection for Deep Sparse Network
This work addresses a specific bottleneck in deep sparse networks for prediction tasks with high-dimensional sparse features, representing an incremental advancement.
The paper tackles the problem of feature interaction selection in deep sparse networks by introducing a hybrid-grained approach that targets both feature field and feature value, resulting in improved accuracy and efficiency as demonstrated on three large real-world benchmark datasets.
Deep sparse networks are widely investigated as a neural network architecture for prediction tasks with high-dimensional sparse features, with which feature interaction selection is a critical component. While previous methods primarily focus on how to search feature interaction in a coarse-grained space, less attention has been given to a finer granularity. In this work, we introduce a hybrid-grained feature interaction selection approach that targets both feature field and feature value for deep sparse networks. To explore such expansive space, we propose a decomposed space which is calculated on the fly. We then develop a selection algorithm called OptFeature, which efficiently selects the feature interaction from both the feature field and the feature value simultaneously. Results from experiments on three large real-world benchmark datasets demonstrate that OptFeature performs well in terms of accuracy and efficiency. Additional studies support the feasibility of our method.