1. Design and develop large-scale causal models to support multi-scenario and multi-objective billion-user-level modeling requirements, and optimize the retention and experience of entertainment live broadcast users 2. Design and develop incentive strategies, depict the long-term impact of strategies on users, quantify the long-term gain of a single incentive for users, optimize delivery and incentive strategies, and design incentive strategies under unbiased data and cost constraints 3. Research breakthroughs in business incentive growth algorithm problems, including large-scale discrete feature deep learning, causal models, relationship networks, operations optimization strategies and other research directions to empower business growth.
1. Have solid coding skills and theoretical foundation of machine learning, and have good coding habits and documentation writing skills 2. Master the theoretical foundation of machine learning, and be familiar with classic algorithm models (GBDT/LR/FM/DNN, etc.) and related tool frameworks (Tensorflow/PyTorch, etc.) 3. Can skillfully use Hive/S estimation, causal inference, Uplift modeling, overall optimization and other projects with practical work experience is preferred 4. Excellent understanding, communication and team collaboration skills, can quickly understand the business background, be sensitive to data, use data facts as the benchmark, and have a strong sense of responsibility.
