Algorithms for Envelope Estimation II

2015/09/12 by Cook, Dennis, Forzani, Liliana, Su, Zhihua
#FOS: Computer and information sciences #Methodology (stat.ME)

paper · doi:10.48550/arxiv.1509.03767

Abstract

We propose a new algorithm for envelope estimation, along with a new root n consistent method for computing starting values. The new algorithm, which does not require optimization over a Grassmannian, is shown by simulation to be much faster and typically more accurate that the best existing algorithm proposed by Cook and Zhang (2015c).

Related

• Algorithms for envelope estimation
• Model-free Envelope Dimension Selection
• Envelope Methods with Ignorable Missing Data
• Envelopes and principal component regression
• Weighted envelope estimation to handle variability in model selection