Mixture models are promising statistical tools aiming to modeling and clustering data arisen from a heterogeneous population. This paper presents a mixture model based on the assumption that the mixing components follow the multivariate restricted skew-normal scale mixture of Birnbaum–Saunders (SNBS) distributions. A computationally feasible expectation-maximization algorithm is developed to carry out maximum likelihood estimation of the new model. Simulation studies are carried out to check the clustering performance and classification accuracy. Finally, illustrative example is presented by analyzing a real-world data set.

中文翻译：

---

伯恩鲍姆-桑德斯分布的多元受限偏斜正态尺度混合物的混合建模

混合模型是有前途的统计工具，旨在对异类种群产生的数据进行建模和聚类。本文基于混合成分遵循伯恩鲍姆-桑德斯（SNBS）分布的多元受限偏斜正态比例混合的假设，提出了一种混合模型。开发了一种在计算上可行的期望最大化算法，以进行新模型的最大似然估计。进行仿真研究以检查聚类性能和分类准确性。最后，通过分析实际数据集给出了说明性示例。