专家简介： 孙玉莹，中国科学院数学与系统科学研究院副研究员。中国科学院大学管理学博士。长期从事时间序列分析、经济预测理论与方法等。2018年以来，学术成果已第一作者发表或接受发表在Journal of Econometrics(ABS4), European Journal of Operational Research(ABS4), Journal of Travel Research(ABS4), Energy Economics, Quantitative Finance, China Economic Review等主流期刊；参与撰写政策研究报告80余篇，其中多篇得到国家领导人批示或被中办、国办采用。曾获得中国科学院数学与系统科学研究院“重要科研进展奖（2017，2019）”、关肇直青年研究奖、中国管理科学与工程学会优秀博士学位论文奖等。
报告摘要：Modeling and forecasting symbolic data, especially interval-valued time series (ITS) data, has received considerable attention in statistics and related fields. The core of available methods on ITS analysis is based on modelling point-valued time series derived from an ITS, which may not efficiently make use of the information contained in interval data. We propose a series of linear and nonlinear models for ITS data by treating an interval as an inseparable entity. By matching the interval models with interval observations, we develop minimum distance estimation methods and establish the asymptotic theory for the proposed estimators. The proposed interval models can be used to address some important issues in economics and finance, e.g., whether are oil stocks efficiently priced in interval-valued factor pricing models? What is the direction and magnitude of Trump Election’s impact on US stock “expected return” and “market efficiency”?
腾讯会议号：801 666 817
时间：2021年12月30日上午10:00