摘要：It is well understood that for conventional survey designs the set of unordered distinct units in a sample is a minimally sufficient statistic. This means that for inferential statistic of the sample, the value of the sampled units rather than the sample design is important. Sampling rare populations presents distinct challenges. Examples of rare populations are in biology with rare and endangered animals where there are only a few remaining individuals, or in social science, with the low incidence of people from an unusually high （or low） income group. Sampling rare populations tends to result in the case that many of the sample units do not contain information on the characteristic of interest （e.g., the rare animal, or people from the unusual income group）. For finite rare populations the set of unordered distinct rare-units in a sample is a minimally sufficient statistic. In an example case study of a rare buttercup, the properties of the minimal sufficient estimator are explored. We compare the efficiency of the estimator for the population total based on the minimally sufficient statistic, with the standard estimator for a range of sample sizes. The variance of the minimally sufficient estimator was always smaller than the variance of the sufficient estimator. For rare populations where non-rare units can be distinguished from rare units because they have the same fixed value, the minimal sufficient statistic is the rare units, if any, in the sample.

摘要：This paper presents a two-level learning method for designing an optimal Radial Basis Function Network (RBFN) using Adaptive Velocity Update Relaxation Particle Swarm Optimization algorithm (AVURPSO) and Orthogonal Least Squares algorithm (OLS) called as OLS-AVURPSO method. The novelty is to develop an AVURPSO algorithm to form the hybrid OLS-AVURPSO method for designing an optimal RBFN. The proposed method at the upper level finds the global optimum of the spread factor parameter using AVURPSO while at the lower level automatically constructs the RBFN using OLS algorithm. Simulation results confirm that the RBFN is superior to Multilayered Perceptron Network (MLPN) in terms of network size and computing time. To demonstrate the effectiveness of proposed OLS-AVURPSO in the design of RBFN, the Mackey-Glass Chaotic Time-Series as an example is modeled by both MLPN and RBFN.