Title: Semiparametric model averaging prediction for dichotomous response
Authors: Fang Fang - East China Normal University (China) [presenting]
Jialiang Li - NUS, Duke-NUS, SERI (Singapore)
Xiaochao Xia - Huazhong Agricultural University (China)
Abstract: Model averaging has attracted abundant attentions from researchers in the past decades as it becomes a powerful forecasting tool in areas such as econometrics, social sciences and medicine. So far, most developed model averaging methods focus only on either parametric models or nonparametric models with a continuous response. We propose a semiparametric model averaging prediction (SMAP) method for a dichotomous response. The idea is to approximate the unknown discrimination score function by a linear combination of one-dimensional marginal score functions. The weight parameters involved in the approximation are constructed by an initial stage nonparametric smoothing estimation of the marginal scores and then applying the familiar parametric model averaging on the dichotomous response based on the likelihood estimation. The proposed model averaging procedure provides more flexibility than parametric models while being more stable than a fully nonparametric approach. Some theoretical properties are investigated and the weight optimality is established under certain technical assumptions. In particular, the weight optimality result requires a condition that reflects the difficulty of model averaging with nonparametric estimates. Empirical evidences from simulation studies and a real data analysis are presented to illustrate our methods.