报告题目：Recent advances in construction of column-orthogonal strong orthogonal arrays and maximin distance designs
报告人：李文龙（北京理工大学）
报告时间：2022年3月4日14:00-16:00
报告地点：腾讯会议号：198-834-989
报告摘要: Strong orthogonal arrays enjoy more attractive space-filling properties than ordinary orthogonal arrays for computer experiments. To further improve the space-filling properties in low dimensions while possessing the column orthogonality, we propose column-orthogonal strong orthogonal arrays of strength two star and three. Construction methods and characterizations of such designs are provided. The resulting strong orthogonal arrays, with the numbers of levels being increased, have their space-filling properties in one and two dimensions being strengthened. They can accommodate comparable or even larger numbers of factors than those in the existing literature, enjoy flexible run sizes, and possess the column orthogonality. In addition to weaken the strength of strong orthogonal arrays while improving the number of factors and possessing the column orthogonality, we propose some methods for constructing column-orthogonal nearly strong orthogonal arrays. These designs enjoy column orthogonality, inherit the nice two-dimensional space-filling property of strong orthogonal arrays, and can accommodate the twice or more number of factors than the existing strong orthogonal arrays. In addition, the proposed designs with four levels enjoy an attractive space-filling property under the maximin distance criterion. The construction methods are convenient and flexible, and the resulting designs are good choices for computer experiments.
One attractive class of space-filling designs for computer experiments is that of maximin distance designs. Algorithmic search for such designs is commonly used but this method becomes ineffective for large problems. Theoretical construction of maximin distance designs is challenging; some results have been obtained recently, often by employing highly specialized techniques. This paper presents an easy-to-use method for constructing maximin distance designs. The method is versatile as it is applicable for any distance measure. Our basic idea is to construct large designs from small designs and the method is effective because the quality of large designs is guaranteed by that of small designs, as evaluated by the maximin distance criterion.
报告人简介：李文龙，北京理工大学数学与统计学院数学博士后，合作导师为田玉斌教授。硕士毕业于华中师范大学概率论与数理统计专业，导师为覃红教授；博士毕业于南开大学统计学专业，导师为刘民千教授；博士期间赴加拿大西蒙弗雷泽大学统计与精算学系联合培养1年，合作导师为Boxin Tang教授。主要研究方向为试验设计和计算机试验。现已在《中国科学》、《Biometrika》、《Journal of Statistical Planning and Inference》、《Statistical Papers》、《Journal of Systems Science and Complexity》等统计学重要期刊上正式接收或发表7篇SCI学术论文，主持中国博士后科学基金面上基金1项。同时，积极参加学术会议并汇报论文成果，曾获“全国第四届统计学博士研究生学术论坛论文评选一等奖”、“第二十三届京津冀青年概率统计学术会议钟家庆优秀论文奖”等。现担任《Statistica Sinica》、《Journal of Statistical Theory and Practice》、《系统科学与数学》期刊的审稿人。
