PPDL: A Real-world Industrial User Retention Ratio Forecasting Framework Integrating Physical Priors with Deep Learning
The proposed PPDL, a novel forecasting framework that integrates physical priors with deep learning, is validated through comprehensive experiments on industrial-scale datasets, and achieves improvements across different backbones and significantly outperforms existing online solutions.
Zi-Bo Zhao, Zheng-Xiong Guan, Chaoli Zhang et al.
· 0 citations