BIVAS: A Scalable Bayesian Method for Bi-Level Variable Selection With Applications

Keyphrases

• Bayesian Methods 100%
• Bi-level Variable Selection 100%
• Numerical Results 25%
• Simulation Study 25%
• Computational Efficiency 25%
• Real-time Data Analysis 25%
• Parameter Estimation 25%
• Computationally Efficient 25%
• Genetic Variants 25%
• Multi-task Learning 25%
• Computationally Intensive 25%
• Gene Level 25%
• Selection Problem 25%
• Mean-field Approximation 25%
• Estimation Efficiency 25%
• Posterior Distribution 25%
• High-dimensional Regression 25%
• Variational Inference 25%
• Group Membership 25%
• Parallelizable Algorithm 25%
• Selection Parameters 25%
• Markov Chain Monte Carlo Methods 25%
• Variational Approximation 25%
• Hierarchical Factorization 25%

Mathematics

• Bayesian 100%
• Simulation Study 50%
• Real Data 50%
• Approximates 50%
• Covariate 50%
• Factorization 50%
• Parameter Estimation 50%
• Markov Chain Monte Carlo Method 50%
• Supplementary Material 50%
• Group Membership 50%
• Variational Approximation 50%
• Posterior Distribution 50%
• Field Approximation 50%