B0599
Title: Inference on some (nearly)-singular covariance matrices
Authors: Karl Oskar Ekvall - University of Florida (United States) [presenting]
Abstract: Recent theory shows reliable inference on small or vanishing variances is possible using connections to inference with singular Fisher information. We briefly review this theory and outline ongoing work on extending the theory to more general settings, including inference on (nearly-)singular covariance matrices. Some computational and practical issues will also be discussed.