Posts by Collection

portfolio

publications

High Dimensional PCA: A New Model Selection Criterion

Published in arXiv preprint, 2020

This paper investigates model selection criteria in high-dimensional PCA, focusing on the Akaike Information Criterion (AIC) and its consistency under varying conditions of eigenvalue separation.

Recommended citation: Abhinav Chakraborty, Soumendu Sundar Mukherjee, Arijit Chakrabarti. (2020). "High Dimensional PCA: A New Model Selection Criterion." arXiv preprint arXiv:2011.04470.
Download Paper

talks