Lower bounds on the smallest eigenvalue of a sample covariance matrix

We derive tight lower bounds on the smallest eigenvalue of a sample covariance matrix of a centred isotropic random vector under weak or no assumptions on its components.