具有未知模式試驗誤差之群組試驗的穩健參數估計法(2/2)

Group testing plays an important role in estimating the prevalence of a rare disease or the proportion of a particular trait. However, test errors often occur and the error rates may vary with the group size. In this project, we will build a new robust prevalence estimator based on γ‐divergence. The minimum γ‐divergence estimator is expected to be consistent without modeling the test error rates, and also reduces the variation by not involving additional parameters in the test error mechanism. We will investigate the asymptotic properties of this prevalence estimation, and study the corresponding optimal design problems. We will also compare the finite sample performance of the obtained designs and that of some commonly-used group testing designs.