Model-based clustering with nonconvex penalty

Abstract

Nonconvex penalty functions, which include the smoothly clipped absolute deviation (SCAD) penalty, minimax concave penalty (MCP) and l q (0 ≤ q 0 and l 1/2 ) to construct specific models. They are compared with the l 1 penalty in the simulations and a real world application. Based on our experiments, the finite sample performance of the four proposed models is well exhibited. In particular, our numerical results suggest that the model-based clustering with the MCP or l 0 penalty is the preferred approach.