Title: Corrected Mallows model averaging approach
Authors: Guohua Zou - School of Mathematical Sciences, Capital Normal University (China) [presenting]
Abstract: An important problem with model averaging approach is the choice of weights. The Mallows criterion for choosing weights is the first asymptotically optimal criterion, which has been used widely. We propose a corrected Mallows model averaging (MMAc) method based on small sample $F$ distribution. MMAc exhibits the same asymptotic optimality as Mallows model averaging (MMA) in the sense of minimizing the squared errors in large sample sizes. The consistency of the MMAc based weights tending to the optimal weights minimizing MSE is also studied. We derive the convergence rate of the new empirical weights. Similar property for MMA and Jackknife model averaging (JMA) is established as well. An extensive simulation study shows that MMAc often performs better than MMA and other commonly used model averaging methods, especially for small and moderate sample size cases. The results from two real data analyses also support the proposed method.